Help improve this workflow!
This workflow has been published but could be further improved with some additional meta data:- Keyword(s) in categories input, output, operation, topic
You can help improve this workflow by suggesting the addition or removal of keywords, suggest changes and report issues, or request to become a maintainer of the Workflow .
Description
GRAPE is an open-source end-to-end Genomic RelAtedness detection PipelinE.
It requires a single multisamples VCF file as input and has a separate workflow for downloading and checking the integrity of reference datasets.
GRAPE incorporates comprehensive data preprocessing, quality control (QC), and several workflows for relatedness inference.
Installation
docker build -t grape:latest -f containers/snakemake/Dockerfile -m 8GB .
The pipeline has three main steps:
-
Downloading of the reference datasets.
-
Quality control and data preprocessing.
-
Application of the relationship inference workflow.
The pipeline is implemented with the
Snakemake framework
and can be accessible through the
snakemake
command.
All the pipeline functionality is embedded into the GRAPE launcher:
launcher.py
, that invokes Snakemake under the hood.
By default, all the commands just check the accessibility of the data and build the computational graph.
To actually perform computation one should add
--real-run
flag to all the commands.
Downloading the Reference Datasets
This step only needs to be done once.
To download reference dataset one should use
reference
command of the pipeline launcher.
The command below downloads needed references to the directory specified by the flag
--ref-directory
.
After that
--ref-directory
argument should be used in all subsequent commands.
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py reference --ref-directory /media/ref --real-run
If phasing and imputation are required (mainly for the GERMLINE workflow) one should specify additional
--phase
and
--impute
flags to previous command to download additional reference datasets.
Total amount of required disk space must be at least 50GB .
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py reference \
--ref-directory /media/ref --phase --impute --real-run
There are another options to download all required reference data as a single file. This file is prepared by GenX team and preloaded on our side in the cloud.
It can be done by specifying additional flag
--use-bundle
to the
reference
command.
This way is faster, since all the post-processing procedures have been already performed.
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py reference --use-bundle \
--ref-directory /media/ref --phase --impute --real-run
Quality Control and Data Preprocessing
GRAPE have a versatile and configurable preprocessing workflow. One part of the preprocessing is required and must be performed before the relationship inference workflow.
Preprocessing is launched by the
preprocess
command of the GRAPE.
Along with some necessary technical procedures, preprocessing includes the following steps.
-
[Required] SNPs quality control by minor allele frequency (MAF) and the missingness rate . We discovered that blocks of rare SNPs with low MAF value in genotype arrays may produce false positive IBD segments. To address this problem, we filter SNPs by minor allele frequency. We remove SNPs with MAF value less than 0.02. Additionally, we remove multiallelic SNPs, insertions / deletions, and SNPs with the high missingness rate, because such SNPs are inconsistent with IBD detection tools.
-
[Required] Per-sample quality control, using missingness and heterozygosity . Extensive testing revealed that samples with an unusually low level of heterozygosity could produce many false relatives matches among individuals. GRAPE excludes such samples from the analysis and creates a report file with the description of the exclusion reason.
-
[Required] Control for strands and SNP IDs mismatches . During this step GRAPE fixes inconsistencies in strands and reference alleles.
-
[Optional] LiftOver from hg38 to hg37 . Currently GRAPE uses hg37 build version of the human genome reference. The pipeline supports input in hg38 and hg37 builds (see
--assembly
flag of the pipeline launcher). If hg38 build is selected (--assembly h38
), then GRAPE applies liftOver tool to the input data in order to match the hg37 reference assembly. -
[Optional] Phasing and imputation . GRAPE supports phasing and genotype imputation using 1000 Genomes Project reference panel. GERMLINE IBD detection tool requires phased data. So, if input data is unphased, one should include phasing (Eagle 2.4.1) (
--phase
flag) into the preprocessing before running the GERMLINE workflow. If input data is highly heterogeneous in a sense of available SNPs positions, one can also add imputation (Minimac4) procedure to the preprocessing (--impute
flag). -
[Optional] Removal of imputed SNPs . We found, that if input data is homogeneous in a sense of SNPs positions, the presence of imputed SNPs does not affect the overall IBD detection accuracy of the IBIS tool, but it significantly slows down the overall performance. For this particular case, when input data initially contains a lot of imputed SNPs, we recommend to remove them by specifying
--remove-imputation
flag to the GRAPE launcher. GRAPE removes all SNPs which are marked withIMPUTED
flag in the input VCF file.
Usages
-
Preprocessing for the IBIS + KING relatedness inference workflow.
-
Input file is located at
/media/input.vcf.gz
. -
GRAPE working directory is
/media/data
. -
Assembly of the input VCF file is
hg37
.
-
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py preprocess --ref-directory /media/ref \
--vcf-file /media/input.vcf.gz --directory /media/data --assembly hg37 --real-run
-
Preprocessing for the GERMLINE + KING workflow.
-
Assembly of the input VCF file is
hg38
. -
GERMLINE can work with phased data only, so we add phasing procedure to the preprocessing.
-
Genotype imputation is also added.
-
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py preprocess --ref-directory /media/ref \
--vcf-file /media/input.vcf.gz --directory /media/data \
--assembly hg38 --phase --impute --flow germline-king --real-run
GRAPE Relatedness Inference Workflows
There are three relationship inference workflows implemented in GRAPE.
These workflows are activated by the
find
command of the launcher.
Workflow selection is made by the
--flow
parameter.
-
IBIS + ERSA ,
--flow ibis
. During this workflow IBD segments detection is performed by IBIS, and estimation of relationship degree is carried out by means of ERSA algorithm. This is the fastest workflow. -
IBIS + ERSA & KING ,
--flow ibis-king
. During this workflow, GRAPE uses KING tool for the first 3 degrees of relationships, and IBIS + ERSA approach for higher order degrees. -
GERMLINE + ERSA & KING ,
--flow germline-king
. The workflow uses GERMLINE for IBD segments detection. KING is used to identify relationship for the first 3 degrees, and ERSA algorithm is used for higher order degrees. This workflow was added to GRAPE mainly for the case, when input data is already phased and accurately preprocessed. -
(BETA!) RaPID + ERSA ,
--flow rapid
This workflow uses RaPID for the IBD segments detection and requires phased and preprocessed data as input.
Usages
- Relationship inference with the IBIS + ERSA & KING workflow.
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py find --flow ibis-king --ref-directory /media/ref \
--directory /media/data --ibis-seg-len 7 --ibis-min-snp 500 \
--zero-seg-count 0.5 --zero-seg-len 5.0 --alpha 0.01 --real-run
- Relationship inference with GERMLINE + ERSA & KING workflow.
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py find --flow germline-king --ref-directory /media/ref \
--directory /media/data --zero-seg-count 0.5 --zero-seg-len 5.0 \
--alpha 0.01 --real-run
Description of the IBIS and ERSA Parameters
-
[IBIS]
--ibis-seg-len
. Minimum length of the IBD segment to be found by IBIS. Higher values reduce false positive rate and give less distant matches (default = 7 cM). -
[IBIS]
--ibis-min-snp
. Minimum number of SNPs per IBD segment to be detected (default = 500 SNPs). -
[ERSA]
--zero-seg-count
. Mean number of shared segments for two unrelated individuals in the population. Smaller values tend to give more distant matches and increase the false positive rate (default = 0.5). -
[ERSA]
--zero-seg-len
. Average length of IBD segment for two unrelated individuals in the population. Smaller values tend to give more distant matches and increase the false positive rate (default = 5 cM). -
[ERSA]
--alpha
. ERSA significance level (default = 0.01).
Description of the RaPID Parameters
-
[RaPID]
--rapid-seg-len
. Minimum length of the IBD segment to be found by RaPID. Higher values reduce false positive rate and give less distant matches (default = 5 cM). -
[RaPID] `--rapid-num-runs``. Number of random projections for RaPID algorithm
-
[RaPID] `--rapid-num-success``. Number of matches for a segment. Higher values should reduce false positive rate but also reduce recall.
Description of the Output Relatives File
Output relatives file is in TSV file format. It contains one line for each detected pair of relatives.
id1 id2 king_degree king_relation shared_genome_proportion kinship kinship_degree ersa_degree ersa_lower_bound ersa_upper_bound shared_ancestors final_degree total_seg_len seg_count
g1-b1-i1 g2-b1-i1 1 PO 0.4996 0.2493 1.0 1 1 1 0.0 1 3359.9362945470302 40
g1-b1-i1 g2-b2-i1 1 PO 0.4997 0.2459 1.0 1 1 1 0.0 1 3362.012253715002 40
g1-b1-i1 g2-b3-i1 1 PO 0.4999 0.2467 1.0 1 1 1 0.0 1 3363.150814131464 40
g1-b1-i1 g3-b1-i1 2 2 0.2369 0.1163 2.0 2 2 2 1.0 2 1644.634182188072 60
-
id1
- ID of the first sample in a pair of relatives. -
id2
- ID of the second sample in a pair of relatives,id1
is always less thanid2
. -
king_degree
- Numeric degree of relationship estimated by KING;-
0
means duplicates or MZ twins, -
1
means parent-offspring (PO), -
2
can be either full siblings (FS), half siblings and grandmother/grandfather with a granddaughter/grandson. -
3
is aunt/uncle with a niece/nephew, as described in table in https://en.wikipedia.org/wiki/Coefficient_of_relationship.
If
king_degree
exists, thenfinal_degree
will be equal toking_degree
. -
-
king_relation
- further differentiation for the first 2 degrees of KING;king_degree
equals1
means PO - parent-offspring, also KING detects FS in some of second degrees. -
shared_genome_proportion
is the approximate fraction of genome shared by two individuals. It should be approximately 0.5 for PO and FS, 0.25 for grandmother-granddaughter and half-siblings. For the first 3 degrees it is calculated as total length of IBD2 segments + half of total length of IBD1 segments. For 4th+ degrees it is simply half of total length of IBD1 segments. -
kinship
is the KING kinship coefficient. -
ersa_degree
is the degree estimated from IBD segments by ERSA, it is used for thefinal_degree
whenking_degree
does not exist. -
ersa_lower_bound
/ersa_upper_bound
is the lower / upper bound of the degree estimated by the ERSA using confidence interval corresponding to the significance level α (--alpha
). Accordingly to the ERSA likelihood model, true degree does not belong to this interval with probability equals to α. -
shared_ancestors
is the most likeliest number of shared ancestors. If it equals0
, then one relative is a direct descendant of the other; if equals1
, then they most probably have one common ancestor and represent half siblings; if equals2
, then, for examples, individuals may have common mother and father. -
final_degree
is the final relationship degree predicted by GRAPE. If KING is used (--flow ibis-king
or--flow germline-king
), it equalsking_degree
for close relatives up to the 3rd degree, andersa_degree
for distant relatives. -
total_seg_len
is the total length of all IBD1 segments. It's calculated using IBIS or GERMLINE IBD data. If KING is involved, then for the first 3 degrees it's calculated using KING IBD data. -
seg_count
is the total number of all IBD segments found by IBIS / GERMLINE tools. If KING is involved, the total number of found IBD segments is taken from KING for the first 3 degrees.
IBD Segments Weighting
Distribution of IBD segments among non-related (background) individuals within a population may be quite heterogeneous. There may exist genome regions with extremely high rates of overall matching, which are not inherited from the recent common ancestors. Instead, these regions more likely reflect other demographic factors of the population.
The implication is that IBD segments detected in such regions are expected to be less useful for estimating recent relationships. Moreover, such regions potentially prone to false-positive IBD segments.
GRAPE provides two options to address this issue.
The first one is based on genome regions exclusion mask, wherein some genome regions are completely excluded from the consideration. This approach is implemented in ERSA and is used by GRAPE by default.
The second one is based on the IBD segments weighing.
The key idea is to down-weight IBD segment, i.e. reduce the IBD segment length, if the segment cross regions with high rate of matching.
Down-weighted segments are then passed to the ERSA algorithm.
GRAPE provides an ability to compute the weight mask from the VCF file with presumably unrelated individuals.
This mask is used during the relatedness detection by specifying the
--weight-mask
flag of the launcher.
See more information in out GRAPE preprint .
Computation of the IBD segments weighing mask
docker run --rm -it -v /media:/media \
grape:latest launcher.py compute-weight-mask \
--directory /media/background --assembly hg37 \
--real-run --ibis-seg-len 5 --ibis-min-snp 400
The resulting files consist of a weight mask file in JSON format and a visualization of the mask stored in
/media/background/weight-mask/
folder.
Usage of the IBD segments weighing mask
docker run --rm -it -v /media:/media \
grape:latest launcher.py find --flow ibis --ref-directory /media/ref \
--weight-mask /media/background/weight-mask/mask.json \
--directory /media/data --assembly hg37 \
--real-run --ibis-seg-len 5 --ibis-min-snp 400
Execution by Scheduler
The pipeline can be executed using lightweight scheduler Funnel , which implements Task Execution Schema developed by GA4GH .
During execution, incoming data for analysis can be obtained in several ways: locally, FTP, HTTPS, S3, Google, etc. The resulting files can be uploaded in the same ways.
It is possible to add other features such as writing to the database, and sending to the REST service.
The scheduler itself can work in various environments from a regular VM to a Kubernetes cluster with resource quotas support.
For more information see the doc .
How to use Funnel:
# Start the Funnel server
/path/to/funnel server run
# Use Funnel as client
/path/to/funnel task create funnel/grape-real-run.json
Launch with Dockstore
GRAPE supports execution with the Dockstore.
Dockstore page for the GRAPE pipeline can be found here .
-
Clone the GRAPE repository.
-
Download reference datasets.
There are four available options:
# Download minimal reference datasets from public available sources (without phasing and imputation)
dockstore tool launch --local-entry workflows/reference/cwl/ref_min.cwl --json workflows/reference/cwl/config.json --script
# Download complete reference datasets from public available sources (for phasing and imputation)
dockstore tool launch --local-entry workflows/reference/cwl/ref.cwl --json workflows/reference/cwl/config.json --script
# Download minimal reference datasets as bundle from (without phasing and imputation)
dockstore tool launch --local-entry workflows/bundle/cwl/bundle_min.cwl --json workflows/reference/bundle/config.json --script
# Download complete reference datasets as bundle from (for phasing and imputation)
dockstore tool launch --local-entry workflows/bundle/cwl/bundle.cwl --json workflows/reference/bundle/config.json --script
- Run the preprocessing.
dockstore tool launch --local-entry workflows/preprocess2/cwl/preprocess.cwl --json workflows/preprocess2/cwl/config.json --script
- Run the relatedness inference workflow.
dockstore tool launch --local-entry workflows/find/cwl/find.cwl --json workflows/find/cwl/config.json --script
Each pipeline step requires
config.json
.
Each config has predefined directories and paths, but you can override these paths by changing them in these files.
Also notice that Dockstore saves its runtime in
datastore
directory.
This directory will grow up with each run.
We recommend to clean it up after each step, especially after reference downloading.
Evaluation on Simulated Data
We added a simulation workflow into GRAPE to perform a precision / recall analysis of the pipeline.
It's accessible by
simulate
command of the pipeline launcher and incorporates the following steps:
-
Pedigree simulation with unrelated founders; we use the Ped-sim simulation package.
-
Relatedness degrees estimation;
-
Comparison between true and estimated degrees.
The source dataset for the simulation is taken from CEU population data of 1000 Genomes Project. As CEU data consists of trios, we picked no more than one member of each trio as a founder.
We also ran GRAPE on selected individuals to remove all cryptic relationships up to the 6th degree.
Then, we randomly assigned sex to each individual and used sex-specific genetic maps to take into account the differences in recombination rates between men and women.
A visualization of structure of simulated pedigree is given below:
How to Run Simulation
Use the
simulate
command of the GRAPE launcher.
docker run --rm -it -v /media:/media -v /etc/localtime:/etc/localtime:ro \
grape:latest launcher.py simulate --flow ibis-king --ref-directory /media/ref \
--directory /media/data --sim-params-file params/relatives_average.def \
--sim-samples-file ceph_unrelated_all.tsv --assembly hg37 --ibis-seg-len 5 \
--ibis-min-snp 400 --zero-seg-count 0.1 --zero-seg-len 5 --alpha 0.01 --real-run
Precision / Recall Analysis
For each
i
degree of relationships we computed precision and recall metrics:
Here TP(i), FP(i), FN(i) are the numbers of true positive, false positive, and false negative relationship matches predicted for the degree
i
.
In our analysis we used non-exact (fuzzy) interval metrics.
-
For the 1st degree, we require an exact match.
-
For the 2nd, 3rd, and 4th degrees, we allow a degree interval of ±1. For example, for the 2nd true degree we consider a predicted 3rd degree as a true positive match.
-
For the 5th+ degrees, we use the ERSA confidence intervals which are typically 3-4 degrees wide.
-
For 10th+ degrees, these intervals are 6-7 degrees wide.
Known Limitations
It is known that for some small isolated populations IBD sharing is very high. Therefore, our pipeline overestimates the relationship degree for them.
It is not recommended to mix standard populations like CEU and the small populations like isolated native ones.
This problem is discussed in https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0034267.
Credits
License: GNU GPL v3
Support
Join our Telegram chat for the support: https://t.me/joinchat/NsX48w4OzcpkNTRi.
Code Snippets
6 7 8 9 | shell: ''' bcftools annotate --set-id "%CHROM:%POS:%REF:%FIRST_ALT" {input.vcf} -O z -o {output.vcf} ''' |
30 31 | script: '../scripts/select_bad_samples.py' |
50 51 | script: '../scripts/plink_filter.py' |
68 69 | script: '../scripts/pre_imputation_check.py' |
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 | shell: """ # remove dublicates cut -f 2 {params.input}.bim | sort | uniq -d > plink/{wildcards.batch}_snp.dups plink --bfile {params.input} --exclude plink/{wildcards.batch}_snp.dups --make-bed --out plink/{wildcards.batch}_merged_filter_dub |& tee -a {log} plink --bfile {params.input} --extract plink/{wildcards.batch}_merged_filter.bim.freq --make-bed --out plink/{wildcards.batch}_merged_freq |& tee -a {log} plink --bfile plink/{wildcards.batch}_merged_freq --extract plink/{wildcards.batch}_merged_filter.bim.chr --make-bed --out plink/{wildcards.batch}_merged_extracted |& tee -a {log} plink --bfile plink/{wildcards.batch}_merged_extracted --flip plink/{wildcards.batch}_merged_filter.bim.flip --make-bed --out plink/{wildcards.batch}_merged_flipped |& tee -a {log} plink --bfile plink/{wildcards.batch}_merged_flipped --update-chr plink/{wildcards.batch}_merged_filter.bim.chr --make-bed --out plink/{wildcards.batch}_merged_chroped |& tee -a {log} plink --bfile plink/{wildcards.batch}_merged_chroped --update-map plink/{wildcards.batch}_merged_filter.bim.pos --make-bed --out {params.out} |& tee -a {log} rm plink/{wildcards.batch}_merged_filter_dub.* rm plink/{wildcards.batch}_merged_freq.* rm plink/{wildcards.batch}_merged_extracted.* rm plink/{wildcards.batch}_merged_flipped.* rm plink/{wildcards.batch}_merged_chroped.* """ |
128 129 130 131 132 133 134 135 136 137 138 139 | shell: """ plink --bfile {params.input} --a1-allele plink/{wildcards.batch}_merged_filter.bim.force_allele --make-bed --out plink/{wildcards.batch}_merged_mapped_alleled |& tee -a {log.plink} plink --bfile plink/{wildcards.batch}_merged_mapped_alleled --keep-allele-order --output-chr M --export vcf bgz --out vcf/{wildcards.batch}_merged_mapped_clean |& tee -a {log.vcf} mkdir vcf/temp_{wildcards.batch} bcftools sort -T vcf/temp_{wildcards.batch} vcf/{wildcards.batch}_merged_mapped_clean.vcf.gz -O z -o {output.temp_vcf} |& tee -a {log.vcf} bcftools index -f {output.temp_vcf} |& tee -a {log.vcf} # need to check output for the potential issues bcftools view {output.temp_vcf} --regions 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22 -O b -o {output.bcf} bcftools norm --check-ref e -f {GRCH37_FASTA} vcf/{wildcards.batch}_merged_mapped_sorted.bcf.gz -O u -o /dev/null |& tee -a {log.vcf} bcftools index -f {output.bcf} | tee -a {log.vcf} """ |
15 16 17 18 19 20 21 22 23 | shell: ''' minimac4 \ --refHaps {input.refHaps} \ --haps phase/{wildcards.batch}_chr{wildcards.chrom}.phased.vcf.gz \ --format GT,GP \ --prefix imputed/{wildcards.batch}_chr{wildcards.chrom}.imputed \ --minRatio 0.01 |& tee {log} ''' |
37 38 39 40 41 42 | shell: ''' FILTER="'strlen(REF)=1 & strlen(ALT)=1'" bcftools view -i 'strlen(REF)=1 & strlen(ALT)=1' imputed/{wildcards.batch}_chr{wildcards.chrom}.imputed.dose.vcf.gz -v snps -m 2 -M 2 -O z -o imputed/{wildcards.batch}_chr{wildcards.chrom}.imputed.dose.pass.vcf.gz |& tee {log} ''' |
71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 | shell: ''' # for now just skip empty files true > {params.list} && \ for i in {input}; do if [ -s $i ] then echo $i >> {params.list} else continue fi done bcftools concat -f {params.list} -O z -o {output} |& tee -a {log} bcftools index -f {output} |& tee -a {log} # check if there is a background data and merge it if [ -f "{params.merged_imputed}" && {params.mode} = "client" ]; then mv {output} {output}.client bcftools merge --force-samples {params.merged_imputed} {output}.client -O z -o {output} |& tee -a {log} bcftools index -f {output} |& tee -a {log} fi ''' |
111 112 113 114 | shell: ''' plink --vcf {input} --make-bed --out {params.out} |& tee {log} ''' |
134 135 136 137 138 139 | shell: ''' # please mind a merge step in merge_imputation_filter for germline plink --vcf {input} --make-bed --out {params.out} | tee {log} plink --bfile {params.background} --bmerge {params.out} --make-bed --out plink/{wildcards.batch}_merged_imputed |& tee {log} ''' |
10 11 12 13 | shell: ''' bcftools index -f -t {input.bcf} |& tee {log} ''' |
25 26 27 28 29 30 31 32 33 34 35 | shell: ''' eagle --vcfRef {input.vcfRef} \ --vcfTarget {input.vcf} \ --geneticMapFile {GENETIC_MAP} \ --chrom {wildcards.chrom} \ --vcfOutFormat b \ --pbwtIters 2 \ --Kpbwt 20000 \ --outPrefix phase/{wildcards.batch}_chr{wildcards.chrom}.phased |& tee {log} ''' |
54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 | shell: ''' # for now just skip empty files true > {params.list} && \ for i in {input}; do if [ -s $i ] then echo $i >> {params.list} else continue fi done bcftools concat -f {params.list} -O z -o {output} |& tee -a {log} bcftools index -f {output} |& tee -a {log} # check if there is a background data and merge it if [ -f "background/{wildcards.batch}_merged_imputed.vcf.gz" && {params.mode} = "client" ]; then mv {output} {output}.client bcftools merge --force-samples background/{wildcards.batch}_merged_imputed.vcf.gz {output}.client -O b -o {output} |& tee -a {log} bcftools index -f {output} |& tee -a {log} fi ''' |
17 18 19 20 21 22 23 24 25 26 27 28 29 | shell: ''' bcftools query --list-samples input.vcf.gz >> vcf/samples.txt total_lines=$(wc -l < vcf/samples.txt) num_files={params.num_batches} ((lines_per_file = (total_lines + num_files - 1) / num_files)) split -l $lines_per_file vcf/samples.txt vcf/batch --additional-suffix=.txt --numeric-suffixes=1 for file in $(find vcf -name 'batch0[1-9].txt') do new=$(echo "$file" | sed "s/0//g") mv "$file" "$new" done ''' |
38 39 40 41 | shell: ''' bcftools view -S {input.samples} {input.vcf} -O z -o {output.vcf} --force-samples ''' |
48 49 50 51 | shell: ''' ln -sr {input.vcf} {output.vcf} ''' |
61 | script: '../scripts/remove_imputation.py' |
68 69 70 71 | shell: ''' ln -sr {input.vcf} {output.vcf} ''' |
86 87 88 89 | shell: ''' JAVA_OPTS="-Xmx{params.mem_gb}g" picard -Xmx12g LiftoverVcf WARN_ON_MISSING_CONTIG=true MAX_RECORDS_IN_RAM=5000 I={input.vcf} O={output.vcf} CHAIN={LIFT_CHAIN} REJECT=vcf/chr{wildcards.batch}_rejected.vcf.gz R={GRCH37_FASTA} |& tee -a {log} ''' |
96 97 98 99 | shell: ''' ln -sr {input.vcf} {output.vcf} ''' |
107 108 109 110 111 112 | shell: ''' bcftools annotate --set-id "%CHROM:%POS:%REF:%FIRST_ALT" {input.vcf} | \ bcftools view -t ^8:10428647-13469693,21:16344186-19375168,10:44555093-53240188,22:16051881-25095451,2:85304243-99558013,1:118434520-153401108,15:20060673-25145260,17:77186666-78417478,15:27115823-30295750,17:59518083-64970531,2:132695025-141442636,16:19393068-24031556,2:192352906-198110229 | \ bcftools view --min-af 0.05 --types snps -O b -o {output.bcf} ''' |
125 126 127 128 | shell: ''' ln -sr {input.bcf} {output.bcf} ''' |
146 147 148 149 | shell: ''' bcftools view {input.bcf} -O z -o {output.vcf} ''' |
167 168 169 170 | shell: ''' plink --vcf {input} --make-bed --out {params.out} |& tee {log} ''' |
192 193 194 195 196 197 198 199 200 201 202 203 204 205 | shell: ''' rm files_list.txt || true for file in {input} do if [[ $file == *.fam ]] then echo ${{file%.*}} >> files_list.txt fi done plink --merge-list files_list.txt --make-bed --out preprocessed/data rm files_list.txt ''' |
213 214 215 216 | shell: ''' bcftools index -f {input.batches_vcf} ''' |
237 238 239 240 241 242 243 244 245 246 247 248 249 | shell: ''' rm complete_vcf_list.txt || true for FILE in {input} do if [[ $FILE == *.gz ]] then echo $FILE >> complete_vcf_list.txt fi done bcftools merge --threads {threads} --file-list complete_vcf_list.txt --force-samples -O z -o {output.vcf} rm complete_vcf_list.txt ''' |
265 266 267 268 | shell: ''' plink --vcf {input} --make-bed --out {params.out} |& tee {log} ''' |
282 283 284 285 | shell: ''' (add-map-plink.pl -cm {input.bim} {params.genetic_map_GRCh37}> {output}) |& tee -a {log} ''' |
292 293 294 295 | shell: ''' bcftools query --list-samples {input.bcf} >> {output.fam} ''' |
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 | shell: """ KING_DEGREE=3 king -b {input.bed} --cpus {threads} --ibdseg --degree $KING_DEGREE --prefix {params.out} |& tee {log} king -b {input.bed} --cpus {threads} --kinship --degree 4 --prefix {params.kin} |& tee -a {log} # we need at least an empty file for the downstream analysis if [ ! -f "{output.king}" ]; then touch {output.king} fi if [ ! -f "{output.kinship}" ]; then touch {output.kinship} fi if [ ! -f "{output.kinship0}" ]; then touch {output.kinship0} fi if [ ! -f "{output.segments}" ]; then touch {output.segments} fi """ |
58 59 60 61 | shell: """ ibis {input.bed} {input.bim} {input.fam} -t {threads} -mt {params.mT} -mL {params.mL} -ibd2 -mL2 3 -hbd -f ibis/merged_ibis |& tee -a {log} """ |
77 78 79 80 81 82 83 | shell: """ python {input.script} \ --input-ibd-segments-file {input.ibd} \ --mask-file {params.mask} \ --output-ibd-segments-file {output.ibd} """ |
97 98 | script: "../scripts/transform_ibis_segments.py" |
122 123 124 125 126 127 | shell: """ zcat {input.ibd} > {output.ibd} ersa --avuncular-adj -ci -a {params.a} --dmax 14 -t {params.t} -l {params.l} \ {params.r} -th {params.th} {output.ibd} -o {output.relatives} |& tee {log} """ |
141 142 143 144 145 146 147 148 149 150 151 | shell: """ FILES="{input}" for input_file in $FILES; do if [[ "$input_file" == "${{FILES[0]}}" ]]; then cat $input_file >> {output} else sed 1d $input_file >> {output} fi done """ |
160 161 | script: "../scripts/split_map.py" |
175 | script: "../scripts/merge_king_ersa.py" |
18 19 20 21 | shell: """ plink --bfile {params.bfile} --chr {params.chr} --make-bed --out ibis_data/{params.chr} --threads {threads} """ |
35 36 37 38 | shell: ''' (add-map-plink.pl -cm {input.bim} {params.genetic_map_GRCh37}> {output}) |& tee -a {log} ''' |
56 57 58 59 | shell: """ ibis {input.bed} {input.bim} {input.fam} -t {threads} -mt {params.mT} -mL {params.mL} -ibd2 -mL2 3 -hbd -gzip -f ibis/merged_ibis_{params.chr} |& tee -a {log} """ |
67 68 69 70 | shell: """ zcat {input} >> {output.ibd} """ |
86 87 88 89 90 91 92 | shell: """ python {input.script} \ --input-ibd-segments-file {input.ibd} \ --mask-file {params.mask} \ --output-ibd-segments-file {output.ibd} """ |
106 107 | script: "../scripts/transform_ibis_segments.py" |
131 132 133 134 135 136 | shell: """ zcat {input.ibd} > {output.ibd} ersa --avuncular-adj -ci -a {params.a} --dmax 14 -t {params.t} -l {params.l} \ {params.r} -th {params.th} {output.ibd} -o {output.relatives} |& tee {log} """ |
150 151 152 153 154 155 156 157 158 159 160 | shell: """ FILES="{input}" for input_file in $FILES; do if [[ "$input_file" == "${{FILES[0]}}" ]]; then cat $input_file >> {output} else sed 1d $input_file >> {output} fi done """ |
171 | script: "../scripts/postprocess_ersa.py" |
16 17 18 19 | shell: """ bcftools index {input} | tee {log} """ |
34 35 36 37 | shell: """ bcftools filter {input} -r {wildcards.chrom} -O z -o vcf/for_rapid_{wildcards.chrom}.vcf.gz | tee {log} """ |
54 55 | script: "../scripts/estimate_rapid_params.py" |
67 68 | script: "../scripts/interpolate.py" |
67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 | script: "../scripts/interpolate.py" ''' rule erase_dosages: input: vcf=rules.index_and_split.output[0] output: vcf='vcf/erased_chr{chrom}.vcf.gz' params: vcf='vcf/erased_chr{chrom}' conda: 'bcftools' shell: "bcftools annotate -x 'fmt' {input.vcf} -O z -o {output.vcf}" ''' |
98 99 100 101 102 | shell: """ rapidibd -i {input.vcf} -g {input.g} -d {params.min_cm_len} -o {params.output_folder}/chr{wildcards.chrom} \ -w {params.window_size} -r {params.num_runs} -s {params.num_success} """ |
110 111 112 113 114 115 116 | shell: """ for file in {input} do zcat ${{file}} >> {output} done """ |
128 129 | script: "../scripts/transform_rapid_segments.py" |
155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 | shell: """ touch {output} FILES="{input.ibd}" TEMPFILE=ersa/temp_relatives.tsv rm -f $TEMPFILE rm -f {output} for input_file in $FILES; do ersa --avuncular-adj -ci -a {params.a} --dmax 14 -t {params.t} -l {params.l} \ {params.r} -th {params.th} $input_file -o $TEMPFILE |& tee {log} if [[ "$input_file" == "${{FILES[0]}}" ]]; then cat $TEMPFILE >> {output} else sed 1d $TEMPFILE >> {output} fi done """ |
185 | script: "../scripts/postprocess_ersa.py" |
19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 | shell: """ KING_DEGREE=3 king -b {params.input}.bed --cpus {threads} --ibdseg --degree $KING_DEGREE --prefix {params.out} |& tee {log} king -b {params.input}.bed --cpus {threads} --kinship --degree $KING_DEGREE --prefix {params.kin} |& tee -a {log} # we need at least an empty file for the downstream analysis if [ ! -f "{output.king}" ]; then touch {output.king} fi if [ ! -f "{output.kinship}" ]; then touch {output.kinship} fi if [ ! -f "{output.kinship0}" ]; then touch {output.kinship0} fi if [ ! -f "{output.segments}" ]; then touch {output.segments} fi """ |
45 | shell: "bcftools index -f {input}" |
59 60 61 62 | shell: """ bcftools filter -r {wildcards.chrom} {input.vcf} | bcftools norm --rm-dup none -O z -o vcf/imputed_chr{wildcards.chrom}.vcf.gz |& tee {log} """ |
80 81 | script: "../scripts/vcf_to_ped.py" |
95 96 97 98 | shell: """ plink --file ped/imputed_chr{wildcards.chrom} --keep-allele-order --cm-map {input.cmmap} {wildcards.chrom} --recode --out cm/chr{wildcards.chrom}.cm |& tee {log} """ |
107 108 109 110 111 112 113 114 115 | shell: """ touch {output} germline -input ped/imputed_chr{wildcards.chrom}.ped cm/chr{wildcards.chrom}.cm.map -min_m 2.5 -err_hom 2 -err_het 1 -output germline/chr{wildcards.chrom}.germline |& tee {log} # TODO: germline returns some length in BP instead of cM - clean up is needed set +e grep -v MB germline/chr{wildcards.chrom}.germline.match > germline/chr{wildcards.chrom}.germline.match.clean && mv germline/chr{wildcards.chrom}.germline.match.clean germline/chr{wildcards.chrom}.germline.match set -e """ |
122 123 | shell: "cat germline/*.match > {output}" |
136 137 | script: '../scripts/merge_ibd.py' |
155 156 157 158 159 | shell: """ touch {output} ersa --avuncular-adj -ci --alpha {params.alpha} --dmax 14 -t {params.ersa_t} -l {params.ersa_l} -th {params.ersa_th} {input.ibd} -o {output} |& tee {log} """ |
175 | script: "../scripts/merge_king_ersa.py" |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 | import scipy import scipy.optimize import random import math import gzip import logging from collections import namedtuple import statsmodels.api as sm lowess = sm.nonparametric.lowess import re import operator as op from functools import reduce import numpy as np w_rho_vals = [] def compute_mafs(vcf_input,max_window_size): global w_rho_vals # w_rho_vals.append(1) f = gzip.open(vcf_input,'rt') entries_started = False expected_vals = [] site_counter = 0 mafs = [] for line in f: if ('#CHROM' in line): entries_started = True elif (entries_started): _values = line.split() _pos = _values[1] alt = _values[4] if (len(alt.split(',')) > 1): continue i = 2 while(i < len(_values) and _values[i] != 'GT'): i += 1 i += 1 tags = _values[7] i = 9 num_ones = 0 num_zeros = 0 while (i < len(_values)): vals = re.split('\||/',_values[i]) if (vals[0] == '1'): num_ones = num_ones + 1 elif (vals[0] == '0'): num_zeros = num_zeros + 1 if (vals[1] == '1'): num_ones = num_ones + 1 elif (vals[1] == '0'): num_zeros = num_zeros + 1 i = i + 1 v = min(num_zeros,num_ones)/float(num_zeros+num_ones) mafs.append(v) f.close() x_vals = [0] y_vals = [0] for o in range(1,max_window_size): window_size = o expected_vals = [] for i in range(0,len(mafs),window_size): expected_maf = 0 _sum = 0.0 for j in range(0,window_size): if (i + j >= len(mafs)): continue _sum += mafs[i+j] for j in range(0,window_size): if (i + j >= len(mafs)): continue if (_sum != 0): expected_maf += (mafs[i+j] * mafs[i+j]/_sum) expected_vals.append(expected_maf) import numpy _a = numpy.array(expected_vals) pa = numpy.percentile(_a,1) rho_v = pa*pa + (1-pa)*(1-pa) x_vals.append(o) y_vals.append(rho_v) w_rho_vals = lowess(y_vals, x_vals,frac=0.1,return_sorted=False) #for k in range(0,len(x_vals)): # print (x_vals[k],w_rho_vals[k]) # for j in w_rho_vals: # print l,rho_v def ncr_old(n, r): r = min(r, n-r) numer = reduce(op.mul, range(n, n-r, -1), 1) denom = reduce(op.mul, range(1, r+1), 1) return numer / denom def w_rho_func(_w): if (_w >= len(w_rho_vals)): return w_rho_vals[len(w_rho_vals)-1] else: # w = int(round(_w,0)) return w_rho_vals[int(_w)] def ncr(n,r): f = math.factorial return f(n) / f(r) / f(n-r) def fp(e,N,L,rho,r,c,w): sum = 0.0 #rho = w_rho_func(w) for i in range(0,c): sum += (ncr(r,i)* ((rho**(L/w))**i) * ( (1-rho**(L/w))**(r-i) )) return 1 - sum def tp(er,N,L,rho,r,c,w): sum = 0.0 for i in range(0,c): sum += (ncr(r,i) * (math.exp(-(er*L)/w)**i) * ((1-math.exp(-(L*er)/w))**(r-i)) ) return 1 - sum def compute_w(error_rate,N,L,rho,r,c,max_w=300): fun = lambda w: 0.5* N*(N-1)*fp(error_rate,N,L,w_rho_func(w),r,c,w) / tp(error_rate,N,L,w_rho_func(int(w)),r,c,w) #bnds = ((0, None)) #x0 = 40 lambda_val = 0.5* N*(N-1) w_min = -1 _started = False for w in range(1,max_w): _tp = tp(error_rate,N,L,w_rho_func(int(w)),r,c,w) _fp = fp(error_rate,N,L,w_rho_func(w),r,c,w) if (round(_tp,2) - 0.5* N*(N-1)* round(_fp,2)) == 1: if (_started): continue else: w_min = w _started = True elif _started: print (w_min, w) return w_min, w return w_min, max_w #cons = ({'type': 'ineq', 'fun': lambda w: 0.5*N*(N-1)*fp(error_rate,N,L,w_rho_func(w),r,c,w) - tp(error_rate,N,L,w_rho_func(int(w)),r,c,w)}) #res = scipy.optimize.minimize(fun,[x0], method='COBYLA', tol=1e-1,bounds=bnds)#,constraints=cons) #print (res) if __name__ == '__main__': try: snakemake except NameError: test_dir = 'test_data/rapid' Snakemake = namedtuple('Snakemake', ['input', 'output', 'params', 'log']) snakemake = Snakemake( input={'vcf': f'test_data/rapid/phased/chr1.phased.vcf.gz', 'num_haplotypes': f'{2504*2}'}, output=[f'{test_dir}/rapid_params'], params={'error_rate': '0.0025', 'min_snps': '600', 'num_runs': '10', 'num_success': '2'}, log=[f'{test_dir}/estimate_rapid_params.log'] ) logging.basicConfig(filename=snakemake.log[0], level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') vcf_input = snakemake.input['vcf'] error_rate = float(snakemake.params['error_rate']) num_haplotypes = int(snakemake.input['num_haplotypes']) min_snps = int(snakemake.params['min_snps']) num_runs = int(snakemake.params['num_runs']) num_success = int(snakemake.params['num_success']) max_window_size = 300 #error_rate = 0.0025 #min_length_SNPs = 12000#14000 rho_initial = 0.9 compute_mafs(vcf_input, max_window_size) print(f'MAFS were computed successfully') w_min, w_max = compute_w(error_rate, num_haplotypes, min_snps, rho_initial, num_runs, num_success, max_window_size) print(f'w_min is {w_min}, w_max is {w_max}') with open(snakemake.output[0], 'w') as out: out.write(f'W={w_min}') |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 | import pandas import numpy from utils.bcftools import bcftools_query def interpolate(vcf_path, cm_path, output_path, chrom): genetic_map = pandas.read_csv(cm_path, sep=" ") pos = list(map(int, bcftools_query(vcf_path, arg="%POS\n", list_samples=False))) # interpolate CM from BP #chro = genetic_map[genetic_map.chr == int(chrom)] # because for chr19 'Genetic_Map(cM)' is 'Genetic_Map.cM.' cms = numpy.interp(pos, genetic_map.position.values, genetic_map.iloc[:, -1].values) eps = 1e-7 new_cms = [] for i in range(0, cms.shape[0]): if i == 0: new_cms.append(cms[i]) continue if new_cms[-1] >= cms[i]: new_cms.append(new_cms[-1] + eps) else: new_cms.append(cms[i]) with open(output_path, 'w') as w_map: for i, cm in enumerate(new_cms): if i > 0 and cm <= new_cms[i-1]: raise ValueError(f'cm positions {i-1} and {i} are NOT strictly increasing: {new_cms[:i+1]}') w_map.write(f"{i}\t{cm:.8f}\n") if __name__ == '__main__': vcf_path = snakemake.input['vcf'] cm_path = snakemake.input['cmmap'] chrom = snakemake.wildcards['chrom'] output_path = snakemake.output[0] interpolate(vcf_path, cm_path, output_path, chrom) |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 | from utils.ibd import merge_germline_segments, segments_to_germline, interpolate_all, read_germline_segments, read_pedsim_segments if __name__ == '__main__': germline_path = snakemake.input['germline'] map_dir = snakemake.params['cm_dir'] gap = float(snakemake.params['merge_gap']) pedsim_path = snakemake.params['true_ibd'] use_true_ibd = bool(snakemake.params['use_true_ibd']) if not use_true_ibd: segments = read_germline_segments(germline_path) segments = interpolate_all(segments, map_dir) segments = merge_germline_segments(segments, gap=gap) segments_to_germline(segments, snakemake.output['ibd']) else: true_segments = read_pedsim_segments(pedsim_path) segments_to_germline(true_segments, snakemake.output['ibd']) |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 | import pandas import numpy import os import logging from collections import Counter, namedtuple from utils.ibd import read_king_segments as rks, interpolate_all, Segment def is_non_zero_file(fpath): return os.path.isfile(fpath) and os.path.getsize(fpath) > 0 def _sort_ids(data): unsorted_mask = data.id2 < data.id1 data.loc[unsorted_mask, 'id1'], data.loc[unsorted_mask, 'id2'] = data.loc[unsorted_mask, 'id2'], data.loc[ unsorted_mask, 'id1'] def read_bucket_dir(bucket_dir): total = None for file in os.listdir(bucket_dir): if not file.endswith('tsv'): continue path = os.path.join(bucket_dir, file) bucket = read_germline(path) if total is None: total = bucket else: total = total.append(bucket) return total def read_germline(ibd_path): germline_names = [ 'fid_iid1', 'fid_iid2', 'chrom', 'start_end_bp', 'start_end_snp', 'snp_len', 'genetic_len', 'len_units', 'mismatches', 'is_homozygous1', 'is_homozygous2' ] data = pandas.read_table(ibd_path, header=None, names=germline_names) if data.shape[0] == 0: return pandas.DataFrame(columns=['id1', 'id2', 'seg_count_germline', 'total_seg_len_germline']).set_index(['id1', 'id2']) data.loc[:, 'id1'] = data.fid_iid1.str.split().str[1] data.loc[:, 'id2'] = data.fid_iid2.str.split().str[1] _sort_ids(data) segments_info = data.loc[:, ['id1', 'id2', 'chrom', 'genetic_len']].groupby(by=['id1', 'id2']).agg({'genetic_len': 'sum', 'chrom': 'count'}) segments_info.rename({'genetic_len': 'total_seg_len_germline', 'chrom': 'seg_count_germline'}, axis=1, inplace=True) return segments_info def map_king_degree(king_degree): degree_map = { 'Dup/MZ': 0, 'PO': 1, 'FS': 2, '2nd': 2, '3rd': 3, '4th': 4, 'NA': numpy.nan } return [degree_map[kd] for kd in king_degree] def map_king_relation(king_degree): degree_map = { 'Dup/MZ': 0, 'PO': 'PO', 'FS': 'FS', '2nd': 2, '3rd': 3, '4th': 4, 'NA': numpy.nan } return [degree_map[kd] for kd in king_degree] def read_king(king_path): # FID1 ID1 FID2 ID2 MaxIBD1 MaxIBD2 IBD1Seg IBD2Seg PropIBD InfType try: data = pandas.read_table(king_path) data.loc[:, 'id1'] = data.FID1.astype(str) + '_' + data.ID1.astype(str) data.loc[:, 'id2'] = data.FID2.astype(str) + '_' + data.ID2.astype(str) _sort_ids(data) data.loc[:, 'king_degree'] = map_king_degree(data.InfType) data.loc[:, 'king_relation'] = map_king_relation(data.InfType) data.loc[:, 'king_degree'] = data.king_degree.astype(float).astype(pandas.Int32Dtype()) data.rename({'PropIBD': 'shared_genome_proportion', 'IBD1Seg': 'king_ibd1_prop', 'IBD2Seg': 'king_ibd2_prop'}, axis='columns', inplace=True) indexed = data.loc[:, ['id1', 'id2', 'king_degree', 'king_relation', 'king_ibd1_prop', 'king_ibd2_prop', 'shared_genome_proportion']].\ set_index(['id1', 'id2']) return indexed except pandas.errors.EmptyDataError: return pandas.DataFrame(data=None, index=None, columns=['id1', 'id2', 'king_degree', 'king_relation', 'king_ibd1_prop', 'king_ibd2_prop', 'shared_genome_proportion']).set_index(['id1', 'id2']) def read_king_segments_chunked(king_segments_path, map_dir): total = None try: for i, chunk in enumerate(pandas.read_table( king_segments_path, compression='gzip', dtype={'FID1': str, 'ID1': str, 'FID2': str, 'ID2': str}, chunksize=1e+6)): segments = {} for i, row in chunk.iterrows(): id1 = row['FID1'] + '_' + row['ID1'] id2 = row['FID2'] + '_' + row['ID2'] seg = Segment(id1, id2, row['Chr'], bp_start=row['StartMB']*1e+6, bp_end=row['StopMB']*1e+6) key = tuple(sorted((seg.id1, seg.id2))) if key not in segments: segments[key] = [seg] else: segments[key].append(seg) segments = interpolate_all(segments, map_dir) data = pandas.DataFrame(columns=['id1', 'id2', 'total_seg_len_king', 'seg_count_king']) logging.info(f'loaded and interpolated segments from chunk {i} for {len(segments)} pairs') rows = [] for key, segs in segments.items(): row = {'id1': key[0], 'id2': key[1], 'total_seg_len_king': sum([s.cm_len for s in segs]), 'seg_count_king': len(segs)} rows.append(row) rows_frame = pandas.DataFrame.from_records(rows, columns=['id1', 'id2', 'total_seg_len_king', 'seg_count_king']) data = pandas.concat([data, rows_frame], ignore_index=True) _sort_ids(data) data = data.set_index(['id1', 'id2']) total = total.append(data) if total is not None else data return total except pandas.errors.EmptyDataError: return pandas.DataFrame(columns=['id1', 'id2', 'total_seg_len_king', 'seg_count_king']).set_index(['id1', 'id2']) def read_king_segments(king_segments_path, map_dir): try: segments = rks(king_segments_path) segments = interpolate_all(segments, map_dir) data = pandas.DataFrame(columns=['id1', 'id2', 'total_seg_len_king', 'seg_count_king']) logging.info(f'loaded and interpolated segments for {len(segments)} pairs') for key, segs in segments.items(): row = {'id1': key[0], 'id2': key[1], 'total_seg_len_king': sum([s.cm_len for s in segs]), 'seg_count_king': len(segs)} data = data.append(row, ignore_index=True) _sort_ids(data) return data.set_index(['id1', 'id2']) except pandas.errors.EmptyDataError: return pandas.DataFrame(columns=['id1', 'id2', 'total_seg_len_king', 'seg_count_king']).set_index(['id1', 'id2']) def kinship_to_degree(kinship): degrees = [] # intervals are from KING manual # >0.354, [0.177, 0.354], [0.0884, 0.177] and [0.0442, 0.0884] for k in kinship: if k > 0.354: degrees.append(0) elif 0.177 < k <= 0.354: degrees.append(1) elif 0.0884 < k <= 0.177: degrees.append(2) elif 0.0442 < k <= 0.0884: degrees.append(3) else: degrees.append(None) return degrees def _read_kinship_data(kinship_path): try: data = pandas.read_table(kinship_path) print(kinship_path, data.shape) # If no relations were found, king creates file with only header if data.shape[0] != 0: if 'FID' in data.columns: data.loc[:, 'id1'] = data.FID.astype(str) + '_' + data.ID1.astype(str) data.loc[:, 'id2'] = data.FID.astype(str) + '_' + data.ID2.astype(str) else: data.loc[:, 'id1'] = data.FID1.astype(str) + '_' + data.ID1.astype(str) data.loc[:, 'id2'] = data.FID2.astype(str) + '_' + data.ID2.astype(str) _sort_ids(data) data.rename({'Kinship': 'kinship'}, axis=1, inplace=True) data.loc[:, 'kinship_degree'] = kinship_to_degree(data.kinship) data = data.loc[:, ['id1', 'id2', 'kinship', 'kinship_degree']].set_index(['id1', 'id2']) else: data = pandas.DataFrame(columns=['id1', 'id2', 'kinship', 'kinship_degree']).set_index(['id1', 'id2']) return data except pandas.errors.EmptyDataError: return pandas.DataFrame(columns=['id1', 'id2', 'kinship', 'kinship_degree']).set_index(['id1', 'id2']) def read_kinship(kinship_path, kinship0_path): # parse within families # FID ID1 ID2 N_SNP Z0 Phi HetHet IBS0 Kinship Error within = _read_kinship_data(kinship_path) logging.info(f'loaded {within.shape[0]} pairs from within-families kinship estimation results') # FID1 ID1 FID2 ID2 N_SNP HetHet IBS0 Kinship across = _read_kinship_data(kinship0_path) logging.info(f'loaded {across.shape[0]} pairs from across-families kinship estimation results') return pandas.concat([within, across], axis=0) def read_ersa(ersa_path): # Indv_1 Indv_2 Rel_est1 Rel_est2 d_est lower_d upper_d N_seg Tot_cM data = pandas.read_table(ersa_path, header=0, names=['id1', 'id2', 'rel_est1', 'rel_est2', 'ersa_degree', 'ersa_lower_bound', 'ersa_upper_bound', 'seg_count_ersa', 'total_seg_len_ersa'], dtype={'ersa_degree': str, 'seg_count_ersa': str, 'total_seg_len_ersa': str}) data = data.loc[(data.rel_est1 != 'NA') | (data.rel_est2 != 'NA'), :] data.loc[:, 'id1'] = data.id1.str.strip() data.loc[:, 'id2'] = data.id2.str.strip() _sort_ids(data) data.loc[:, 'ersa_degree'] = pandas.to_numeric(data.ersa_degree.str.strip(), errors='coerce').astype( pandas.Int32Dtype()) data.loc[:, 'ersa_lower_bound'] = pandas.to_numeric(data.ersa_lower_bound.str.strip(), errors='coerce').astype( pandas.Int32Dtype()) data.loc[:, 'ersa_upper_bound'] = pandas.to_numeric(data.ersa_upper_bound.str.strip(), errors='coerce').astype( pandas.Int32Dtype()) print('dtypes are: ', data.dtypes) data.loc[:, 'seg_count_ersa'] = pandas.to_numeric(data.seg_count_ersa.str.strip(), errors='coerce').astype( pandas.Int32Dtype()) data.loc[:, 'total_seg_len_ersa'] = pandas.to_numeric(data.total_seg_len_ersa.str.strip(), errors='coerce') data.loc[data.ersa_lower_bound != 1, 'ersa_lower_bound'] = data.ersa_lower_bound - 1 data.loc[data.ersa_upper_bound != 1, 'ersa_upper_bound'] = data.ersa_upper_bound - 1 logging.info(f'read {data.shape[0]} pairs from ersa output') logging.info(f'ersa after reading has {pandas.notna(data.ersa_degree).sum()}') data.loc[:, 'is_niece_aunt'] = [True if 'Niece' in d else False for d in data.rel_est1] logging.info(f'we have total of {data.is_niece_aunt.sum()} possible Niece/Aunt relationships') return data.loc[data.id1 != data.id2, ['id1', 'id2', 'ersa_degree', 'ersa_lower_bound', 'ersa_upper_bound', 'is_niece_aunt', 'seg_count_ersa', 'total_seg_len_ersa']].\ set_index(['id1', 'id2']) def read_ersa2(ersa_path): # individual_1 individual_2 est_number_of_shared_ancestors est_degree_of_relatedness 0.95 CI_2p_lower:4 # 2p_upper:5 1p_lower:6 1p_upper:7 0p_lower:8 0p_upper:9 # maxlnl_relatedness maxlnl_unrelatednes data = pandas.read_table(ersa_path, comment='#') data = data.loc[data.est_degree_of_relatedness != 'no_sig_rel', :] data.rename({'individual_1': 'id1', 'individual_2': 'id2', 'est_number_of_shared_ancestors': 'shared_ancestors'}, axis=1, inplace=True) data.loc[:, 'ersa_degree'] = pandas.to_numeric(data['est_degree_of_relatedness'], errors='coerce').\ astype(pandas.Int32Dtype()) lower, upper = [], [] for i, ancestors in enumerate(data.shared_ancestors): if ancestors == 0: lower.append(data.iloc[i, 8]) upper.append(data.iloc[i, 9]) if ancestors == 1: lower.append(data.iloc[i, 6]) upper.append(data.iloc[i, 7]) if ancestors == 2: lower.append(data.iloc[i, 4]) upper.append(data.iloc[i, 5]) data.loc[:, 'ersa_lower_bound'] = lower data.loc[:, 'ersa_upper_bound'] = upper cols = ['id1', 'id2', 'ersa_degree', 'ersa_lower_bound', 'ersa_upper_bound', 'shared_ancestors'] return data.loc[data.id1 != data.id2, cols].set_index(['id1', 'id2']) if __name__ == '__main__': try: snakemake except NameError: test_dir = '/media_ssd/pipeline_data/TF-CEU-TRIBES-ibis-king-2' Snakemake = namedtuple('Snakemake', ['input', 'output', 'params', 'log']) snakemake = Snakemake( input={'king': f'{test_dir}/king/data.seg', 'king_segments': f'{test_dir}/king/data.segments.gz', 'kinship': f'{test_dir}/king/data.kin', 'kinship0': f'{test_dir}/king/data.kin0', 'ersa': f'{test_dir}/ersa/relatives.tsv'}, output=[f'test_data/merge_king_ersa.tsv'], params={'cm_dir': f'{test_dir}/cm'}, log=['test_data/merge_king_ersa.log'] ) logging.basicConfig(filename=snakemake.log[0], level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') king_path = snakemake.input['king'] king_segments_path = snakemake.input['king_segments'] # within families kinship_path = snakemake.input['kinship'] # across families kinship0_path = snakemake.input['kinship0'] ersa_path = snakemake.input['ersa'] map_dir = snakemake.params['cm_dir'] output_path = snakemake.output[0] king = read_king(king_path) kinship = read_kinship(kinship_path, kinship0_path) king_segments = read_king_segments_chunked(king_segments_path, map_dir) ersa = read_ersa(ersa_path) relatives = king.merge(kinship, how='outer', left_index=True, right_index=True).\ merge(ersa, how='outer', left_index=True, right_index=True).\ merge(king_segments, how='outer', left_index=True, right_index=True) prefer_ersa_mask = pandas.isnull(relatives.king_degree) | (relatives.king_degree > 3) relatives.loc[:, 'final_degree'] = relatives.king_degree # if king is unsure or king degree > 3 then we use ersa distant relatives estimation relatives.loc[prefer_ersa_mask, 'final_degree'] = relatives.ersa_degree logging.info(f'king is null or more than 3: {prefer_ersa_mask.sum()}') logging.info(f'ersa is not null: {pandas.notna(relatives.ersa_degree).sum()}') print(f'relatives columns are {relatives.columns}') if 'total_seg_len_king' in relatives.columns: relatives.loc[:, 'total_seg_len'] = relatives.total_seg_len_king*relatives.king_ibd1_prop relatives.loc[:, 'total_seg_len_ibd2'] = relatives.total_seg_len_king*relatives.king_ibd2_prop relatives.loc[:, 'seg_count'] = relatives.seg_count_king relatives.loc[prefer_ersa_mask, 'total_seg_len'] = relatives.total_seg_len_ersa relatives.loc[prefer_ersa_mask, 'seg_count'] = relatives.seg_count_ersa print('is na: ', pandas.isna(relatives.loc[prefer_ersa_mask, 'total_seg_len_ersa']).sum()) # approximate calculations, IBD share is really small in this case relatives.loc[prefer_ersa_mask, 'shared_genome_proportion'] = 0.5*relatives.loc[prefer_ersa_mask, 'total_seg_len'].values / 3440 relatives.drop(['total_seg_len_king', 'seg_count_king', 'total_seg_len_ersa', 'seg_count_ersa', 'king_ibd1_prop', 'king_ibd2_prop'], axis='columns', inplace=True) logging.info(f'final degree not null: {pandas.notna(relatives.final_degree).sum()}') relatives.loc[pandas.notna(relatives.final_degree), :].to_csv(output_path, sep='\t') |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 | import subprocess import logging if __name__ == '__main__': vcf = snakemake.input['vcf'] bad_samples = snakemake.input['bad_samples'] out = snakemake.params['out'] batch = snakemake.params['batch'] log = snakemake.log[0] logging.basicConfig(filename=log, level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') p_freqx = subprocess.run(f'plink --vcf {vcf} --freqx --out plink/{out}', shell=True, capture_output=True) p_remove = subprocess.run(f'plink --vcf {vcf} --remove {bad_samples} ' f'--make-bed --keep-allele-order --out plink/{out}', shell=True, capture_output=True) stdout_freqx = p_freqx.stdout.decode() stderr_freqx = p_freqx.stderr.decode() stderr_remove = p_remove.stderr.decode() stdout_remove = p_remove.stdout.decode() empty_batch_err_code = 11 logging.info(f'\nSTDOUT:\n{stdout_remove}\n{stdout_freqx}' f'\nSTDERR:\n{stderr_remove}\n{stderr_freqx}') if p_remove.returncode != 0 or p_freqx.returncode != 0: if p_remove.returncode == empty_batch_err_code: with open('pass_batches.list', 'r') as list: lines = list.readlines() with open('pass_batches.list', 'w') as list: for line in lines: if line.strip('\n') != f'{batch}': list.write(line) else: print(f'Removed {line}!') else: raise Error(f"Rule plink_filter for batch{batch} failed with error! See {log} for details.") |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 | import polars as pl import logging def _sort_ids(data): return data.with_columns( pl.when(pl.col('id2') < pl.col('id1')) .then( pl.struct([ pl.col('id2').alias('id1'), pl.col('id1').alias('id2') ]) ) .otherwise( pl.struct([ pl.col('id1').alias('id1'), pl.col('id2').alias('id2') ]) ) ).drop('id2').unnest('id1') def read_ibis(ibd_path): # sample1 sample2 chrom phys_start_pos phys_end_pos IBD_type # genetic_start_pos genetic_end_pos genetic_seg_length marker_count error_count error_density def get_new_col_names(old_names): return ['id1', 'id2', 'chrom', 'phys_start_pos', 'phys_end_pos', 'IBD_type', 'genetic_start_pos', 'genetic_end_pos', 'genetic_seg_length', 'marker_count', 'error_count', 'error_density'] data = pl.scan_csv(ibd_path, has_header=False, with_column_names=get_new_col_names, separator='\t', null_values="NA").lazy() data = data.with_columns([ pl.col('id1').str.replace(':', '_'), pl.col('id2').str.replace(':', '_') ]) data = _sort_ids(data) ibd2_info = ( data .filter((pl.col('IBD_type') == 'IBD2')) .select(['id1', 'id2', 'chrom', 'genetic_seg_length']) .groupby(['id1', 'id2']) .agg([ pl.col('chrom').count(), pl.col('genetic_seg_length').sum() ]) ) ibd2_info = ibd2_info.rename({'genetic_seg_length': 'total_seg_len_ibd2', 'chrom': 'seg_count_ibd2'}) return ibd2_info def read_ersa(ersa_path): # Indv_1 Indv_2 Rel_est1 Rel_est2 d_est lower_d upper_d N_seg Tot_cM def get_new_col_names(old_names): return ['id1', 'id2', 'rel_est1', 'rel_est2', 'ersa_degree', 'ersa_lower_bound', 'ersa_upper_bound', 'seg_count', 'total_seg_len'] data = pl.scan_csv(ersa_path, has_header=True, separator='\t', null_values="NA", with_column_names=get_new_col_names, dtypes={'ersa_degree': str, 'ersa_lower_bound': str, 'ersa_upper_bound': str, 'seg_count': str, 'total_seg_len': str}) data = data.filter((pl.col('rel_est1').is_null().is_not()) | (pl.col('rel_est2').is_null().is_not())) data = data.with_columns([ pl.col('id1').str.strip(), pl.col('id2').str.strip() ]) data = _sort_ids(data) print(list(zip(data.columns, data.dtypes))) data = data.with_columns([ pl.col('ersa_degree').str.strip().cast(pl.Int32, strict=False), pl.col('ersa_lower_bound').str.strip().cast(pl.Int32, strict=False), pl.col('ersa_upper_bound').str.strip().cast(pl.Int32, strict=False), pl.col('seg_count').str.strip().cast(pl.Int32, strict=False), pl.col('total_seg_len').str.replace(',', '').str.strip().cast(pl.Float64, strict=False) ]) found = data.select([pl.col('total_seg_len').where(pl.col('total_seg_len') > 1000).sum()]) # print(f'parsed total_seg_len, found {found} long entries') # print(f'{data.filter(pl.col("ersa_lower_bound").is_null()).collect(streaming=True)}') # print(data.with_columns(pl.col('ersa_lower_bound') - 1).collect(streaming=True).head(5)) data = data.with_columns([ pl.when(pl.col('ersa_lower_bound') != 1) .then(pl.col('ersa_lower_bound') - 1).otherwise(pl.col('ersa_lower_bound')), pl.when(pl.col('ersa_upper_bound') != 1) .then(pl.col('ersa_upper_bound') - 1).otherwise(pl.col('ersa_upper_bound')) ]) print(data) print(data.collect(streaming=True).head(5)) #logging.info(f'read {data.shape[0]} pairs from ersa output') #logging.info(f'ersa after reading has {len(data) - data["ersa_degree"].null_count()}') data = data.select( ['id1', 'id2', 'ersa_degree', 'ersa_lower_bound', 'ersa_upper_bound', 'seg_count', 'total_seg_len']) \ .filter(pl.col('id1') != pl.col('id2')) # Removed .set_index(['id1', 'id2']) return data if __name__ == '__main__': logging.basicConfig(filename='log.log', level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') ibd_path = snakemake.input['ibd'] # within families # across families ersa_path = snakemake.input['ersa'] output_path = snakemake.output[0] with open(ersa_path, 'r') as ersa_file, open(ibd_path, 'rb') as ibd_file: if len(ersa_file.readlines(5000)) < 2 or len(ibd_file.readlines(5000)) < 1: print("ersa postprocess input is empty") open(output_path, "w").close() # create empty output to avoid error quit() if snakemake.params.get('ibis', True): ibd = read_ibis(ibd_path) else: _columns = [ ('id1', pl.Utf8), ('id2', pl.Utf8), ('total_seg_len_ibd2', pl.Float64), ('seg_count_ibd2', pl.Int32) ] ibd = pl.DataFrame(data=[], schema=_columns).lazy() ersa = read_ersa(ersa_path) #logging.info(f'ibd shape: {ibd.shape[0]}, ersa shape: {ersa.shape[0]}') relatives = ibd.join(ersa, how='outer', on=['id1', 'id2']) # No left_index / right_index input params # It is impossible to have more than 50% of ibd2 segments unless you are monozygotic twins or duplicates. GENOME_CM_LEN = 3400 DUPLICATES_THRESHOLD = 1750 FS_TOTAL_THRESHOLD = 2100 FS_IBD2_THRESHOLD = 450 dup_mask = pl.col('total_seg_len_ibd2') > DUPLICATES_THRESHOLD fs_mask = (pl.col('total_seg_len') > FS_TOTAL_THRESHOLD) & (pl.col('total_seg_len_ibd2') > FS_IBD2_THRESHOLD) & ( dup_mask.is_not()) po_mask = (pl.col('total_seg_len') / GENOME_CM_LEN > 0.77) & (fs_mask.is_not()) & (dup_mask.is_not()) relatives = relatives.with_columns([ pl.when(dup_mask) .then(0) .when(fs_mask) .then(2) .when(po_mask) .then(1) #.when((pl.col('ersa_degree') == 1)) # when ersa_degree is 1 but PO mask does not tell us that it is PO #.then(2) .otherwise(pl.col('ersa_degree')) .alias('final_degree'), pl.when(dup_mask) .then('MZ/Dup') .when(fs_mask) .then('FS') .when(po_mask) .then('PO') #.when((pl.col('ersa_degree') == 1)) # when ersa_degree is 1 but PO mask does not tell us that it is PO #.then('FS') .otherwise(pl.col('ersa_degree')) .alias('relation'), pl.col('total_seg_len_ibd2') .fill_null(pl.lit(0)), pl.col('seg_count_ibd2') .fill_null(pl.lit(0)) ]) ''' IBIS and ERSA do not distinguish IBD1 and IBD2 segments shared_genome_proportion is a proportion of identical alleles i.e. shared_genome_proportion of [(0/0), (0/1)] and [(0/0), (1/1)] should be 3/4 GENOME_SEG_LEN is length of centimorgans of the haplotype ibd2 segments span two haplotypes, therefore their length should be doubled total_seg_len includes both ibd1 and ibd2 segments length ''' shared_genome_formula = (pl.col('total_seg_len') - pl.col('total_seg_len_ibd2')) / (2*GENOME_CM_LEN) + pl.col('total_seg_len_ibd2')*2 / (2*GENOME_CM_LEN) relatives = relatives.with_columns( pl.when(shared_genome_formula > 1) .then(1) .otherwise(shared_genome_formula) .alias('shared_genome_proportion') ) relatives = relatives.filter(~(pl.col('final_degree').is_null())).collect(streaming=True) print(f'We have {len(relatives.filter(dup_mask))} duplicates, ' f'{len(relatives.filter(fs_mask))} full siblings and ' f'{len(relatives.filter(po_mask))} parent-offspring relationships') logging.info(f'final degree not null: {len(relatives["final_degree"]) - relatives["final_degree"].null_count()}') relatives.write_csv(output_path, separator='\t') |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 | import logging import re ATGC = {'A': 'T', 'T': 'A', 'C': 'G', 'G': 'C'} atgc = {'A': 1, 'T': 1, 'G': 0, 'C': 0} def is_ambiguous(a1, a2): """if SNP is ambiguous or not""" if atgc[a1] == atgc[a2]: return True return False def match_ref(a1, a2, ref, alt): """Match a1, a2 with ref and alt None: match, 1: swap, 2: flip, 3: flip&swap, 4: exclude """ if a1 == ref and a2 == alt: return elif a1 == alt and a2 == ref: return 1 else: ambiguous = is_ambiguous(a1, a2) if ambiguous: return 4 else: a1_f, a2_f = [ATGC[i] for i in [a1, a2]] if a1_f == ref and a2_f == alt: return 2 elif a1_f == alt and a2_f == ref: return 3 else: return 4 def get_id(a1_a2, id_element, chr_, pos_, ref_, alt_): # first, try to use rs id_ = id_element.split(':')[0] if id_ in a1_a2: return id_ # then, try chr:pos id_ = "{}:{}".format(chr_, pos_) if id_ in a1_a2: return id_ # try chr:pos:ref:alt id_ = "{}:{}:{}:{}".format(chr_, pos_, ref_, alt_) if id_ in a1_a2: return id_ # try chr:pos:alt:ref id_ = "{}:{}:{}:{}".format(chr_, pos_, alt_, ref_) if id_ in a1_a2: return id_ return None def pre_imputation_check(params, reference): """Given a plink bim file 0. Remove SNPS with worldwide MAF < 0.02 1. Remove SNPs can not be matched 2. Flip SNPs that match after flip 3. Swap SNPs that match after swap or flip&swap 4. Update chromosomes and positions for these SNPs """ prefix = params fn_new = prefix a1_a2 = {} for i in open(fn_new): items = i.split() # id: (ref, alt) # %CHROM:%POS:%REF:%FIRST_ALT a1_a2[items[1]] = (items[-2], items[-1]) # files to update bim chr_path = prefix + '.chr' pos_path = prefix + '.pos' force_allele_path = prefix + '.force_allele' flip_path = prefix + '.flip' freq_path = prefix + '.freq' with open(chr_path, 'w') as chr_file, open(pos_path, 'w') as pos_file, open(force_allele_path, 'w') as force_allele_file, open(flip_path, 'w') as flip_file, open(freq_path, 'w') as freq_file: af_regex = re.compile(';AF=(0.\d+)') # write the files to be used in_ref = 0 exclude = 0 strand_flip = 0 swap = 0 flip_swap = 0 unique_ids = set() for i in open(reference): items = i.split() str_maf = af_regex.findall(items[7]) maf = float(str_maf[0]) if len(str_maf) > 0 else 0.0 chr_, pos_, ref_, alt_ = items[0], items[1], items[3], items[4] # first, try to use rss id_ = get_id(a1_a2, items[2], chr_, pos_, ref_, alt_) if id_ is not None: a1, a2 = a1_a2[id_] matching = match_ref(a1, a2, ref_, alt_) in_ref += 1 if matching == 4: exclude += 1 else: if id_ in unique_ids: continue if maf > 0.02: freq_file.write(id_ + "\n") chr_file.write("{}\t{}\n".format(id_, chr_)) pos_file.write("{}\t{}\n".format(id_, pos_)) # set alt as A1, because recode vcf will code A1 as alt later force_allele_file.write("{}\t{}\n".format(id_, alt_)) unique_ids.add(id_) if matching == 2: flip_file.write(id_ + "\n") strand_flip += 1 elif matching == 1: swap += 1 elif matching == 3: flip_file.write(id_ + "\n") flip_swap += 1 logging.info("Exclude: {} Keep: {}".format(exclude, in_ref - exclude)) logging.info("Total flip: {}.".format(strand_flip)) logging.info("Total swap: {}.".format(swap)) logging.info("Total flip&swap: {}.".format(flip_swap)) return flip_path, force_allele_path, chr_path, pos_path if __name__ == "__main__": logging.basicConfig(filename=snakemake.log[0], level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') pre_imputation_check(snakemake.input[0], snakemake.params[0]) |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 | import gzip import logging if __name__ == '__main__': genotyped_count = 0 input_file = snakemake.input['vcf'] output_file = snakemake.output['vcf'] logging.basicConfig(filename=snakemake.log[0], level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') with gzip.open(output_file, 'ab') as gzab: with gzip.open(input_file, 'r') as gzr: for cnt, line in enumerate(gzr): if (cnt + 1) % 100000 == 0: logging.info(f'processed {cnt} variants') if b'IMPUTED' not in line: genotyped_count += 1 gzab.write(line) logging.info(f'processed {cnt} variants, removed {cnt - genotyped_count} variants, {genotyped_count} variants remained') |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 | import pandas import logging from utils.bcftools import bcftools_stats from utils.bcftools import bcftools_query from utils.bcftools import bcftools_view from io import StringIO def find_outliers(psc: pandas.DataFrame, outliers_file_path: str, keep_samples_file_path: str, alpha: float): q1 = psc.nNonMissing.quantile(0.25) q3 = psc.nNonMissing.quantile(0.75) iqr = q3-q1 lower_bound_mask = psc.nNonMissing < (q1 - alpha*iqr) upper_bound_mask = psc.nNonMissing > (q3 + alpha*iqr) outliers = psc[lower_bound_mask | upper_bound_mask] if outliers.empty: outliers['exclusion_reason'] = [] else: outliers.loc[lower_bound_mask, 'exclusion_reason'] = f'Sample was below 1st quantile - IQR*{alpha} ({q1 - alpha*iqr}) by Missing Samples' outliers.loc[upper_bound_mask, 'exclusion_reason'] = f'Sample was above 3rd quantile + IQR*{alpha} ({q3 + alpha*iqr}) by Missing Samples' keep_samples = pandas.concat([psc, outliers, outliers]).drop_duplicates(keep=False) outliers_list = list(outliers.sample_id) with open(outliers_file_path, 'w') as outliers_file: for outlier in outliers_list: outliers_file.write(outlier + '\n') keep_samples_list = list(keep_samples.sample_id) with open(keep_samples_file_path, 'w') as keep_samples_file: for sample in keep_samples_list: keep_samples_file.write(sample + '\n') return outliers def get_stats(vcf_input, samples_path, psc_path=False, stats_path=''): bcftools_query(vcf_input, list_samples=True, save_output=samples_path) raw_table = bcftools_stats(vcf_input, samples_path, save_output=psc_path, save_stats=stats_path) names = [ 'PSC', 'id', 'sample_id', 'nRefHom', 'nNonRefHom', 'nHets', 'nTransitions', 'nTransversions', 'nIndels', 'average depth', 'nSingletons', 'nHapRef', 'nHapAlt', 'nMissing' ] psc = pandas.read_table(StringIO(raw_table), header=None, names=names) psc.loc[:, 'nNonMissing'] = psc.nRefHom + psc.nNonRefHom + psc.nHets psc.loc[:, 'missing_share'] = psc.nMissing / (psc.nMissing + psc.nNonMissing) psc.loc[:, 'alt_hom_share'] = psc.nNonRefHom / psc.nNonMissing psc.loc[:, 'het_samples_share'] = psc.nHets / psc.nNonMissing return psc if __name__ == '__main__': input_vcf = snakemake.input['vcf'] stats_file = snakemake.output['stats'] samples = snakemake.params['samples'] psc_path = snakemake.params['psc'] keep_samples = snakemake.params['keep_samples'] bad_samples_path = snakemake.output['bad_samples'] report_path = snakemake.output['report'] outliers = snakemake.output['outliers'] no_outliers_vcf = snakemake.output['no_outliers_vcf'] missing_samples = float(snakemake.params['missing_samples']) alt_hom_samples = float(snakemake.params['alt_hom_samples']) het_samples = float(snakemake.params['het_samples']) alpha = float(snakemake.params['iqr_alpha']) logging.basicConfig(filename=snakemake.log[0], level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') # get initial stats, that we do not save logging.info("Getting initial vcf stats...") psc = get_stats(vcf_input=input_vcf, samples_path=samples, psc_path=psc_path, stats_path=stats_file) # filter outliers outliers = find_outliers(psc, outliers_file_path=outliers, keep_samples_file_path=keep_samples, alpha=alpha) logging.info(f"Found {len(outliers)} outliers!") if not outliers.empty: bcftools_view(input_vcf, no_outliers_vcf, keep_samples) else: with open(no_outliers_vcf, 'w') as dummy_no_outliers_vcf: dummy_no_outliers_vcf.write('') no_outliers_vcf = input_vcf # get final stats without outliers logging.info("Getting final vcf stats...") psc = get_stats(vcf_input=no_outliers_vcf, samples_path=samples, psc_path=psc_path, stats_path=stats_file) bad_missing_samples_mask = (psc.missing_share >= missing_samples / 100) | (psc.nMissing + psc.nNonMissing == 0) bad_alt_hom_samples_mask = (psc.alt_hom_share < alt_hom_samples / 100) | (psc.nNonMissing == 0) bad_het_samples_mask = (psc.het_samples_share < het_samples / 100) | (psc.nNonMissing == 0) psc.loc[bad_het_samples_mask, 'exclusion_reason'] = f'Sample has <= {het_samples}% heterozygous SNPs' psc.loc[bad_alt_hom_samples_mask, 'exclusion_reason'] = f'Sample has <= {alt_hom_samples}% homozygous alternative SNPs' psc.loc[bad_missing_samples_mask, 'exclusion_reason'] = f'Sample has >= {missing_samples}% missing SNPs' bad_samples = psc.loc[(bad_alt_hom_samples_mask | bad_het_samples_mask | bad_missing_samples_mask), ['sample_id', 'missing_share', 'alt_hom_share', 'het_samples_share', 'exclusion_reason']] psc = pandas.concat([psc, outliers[['sample_id', 'missing_share', 'alt_hom_share', 'het_samples_share', 'exclusion_reason']]], ignore_index=True) samples_only = bad_samples.loc[:, ['sample_id']].copy() # if input vcf file has iids in form fid_iid we split it, else we just assign fid equal to iid samples_only.loc[:, 'fid'] = [s.split('_')[0] if '_' in s else s for s in samples_only.sample_id] samples_only.loc[:, 'iid'] = [s.split('_')[1] if '_' in s else s for s in samples_only.sample_id] samples_only.loc[:, ['fid', 'iid']].to_csv(bad_samples_path, sep='\t', header=None, index=False) bad_samples.to_csv(report_path, sep='\t') log = f'We have total of {psc.shape[0]} samples\n' \ f'{bad_missing_samples_mask.sum()} samples have >= {missing_samples}% missing share\n' \ f'{bad_alt_hom_samples_mask.sum()} samples have <= {alt_hom_samples}% homozygous alternative variants\n' \ f'{bad_het_samples_mask.sum()} samples have <= {het_samples}% heterozygous variants\n' \ f'{len(outliers)} samples detected as outliers\n' logging.info(log) print(log) |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 | import pandas import os if __name__ == '__main__': path = snakemake.input['bim'] map_dir = snakemake.params['cm_dir'] bim = pandas.read_table(path, header=None, names=['chrom', '_id', 'cm_pos', 'bp_pos', 'ref', 'alt']) maps = bim.loc[:, ['chrom', '_id', 'cm_pos', 'bp_pos']].groupby(by='chrom') for chrom, cmmap in maps: map_path = os.path.join(map_dir, f'chr{chrom}.cm.map') frame = cmmap.reset_index() frame.to_csv(map_path, index=False, sep='\t', header=None) print(f'written map for chrom {chrom} to {map_path}') print(f'written {len(maps)} maps to {map_dir}') |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 | import pandas import logging import mmh3 from typing import List import os from collections import namedtuple def process_chunk_with_hash(data: pandas.DataFrame, denominator: int, dest_dir: str): data.loc[:, 'id1'] = data.sample1.str.replace(':', '_') + ' ' + data.sample1.str.replace(':', '_') data.loc[:, 'id2'] = data.sample2.str.replace(':', '_') + ' ' + data.sample2.str.replace(':', '_') # awk '{{sub(":", "_", $1); sub(":", "_", $2); print $1, $1 "\t" $2, $2 "\t" $3 "\t" $4, $5 "\t" 0, 0 "\t" $10 "\t" $9 "\t" "cM" "\t" 0 "\t" 0 "\t" 0}}; data.loc[:, 'phys_pos'] = [f'{start} {end}' for start, end in zip(data.phys_start_pos, data.phys_end_pos)] data.loc[:, 'zero1'] = '0 0' data.loc[:, 'zero2'] = '0' data.loc[:, 'cM'] = 'cM' data.loc[:, 'ibd21'] = [2 if ibd_type == 'IBD2' else 0 for ibd_type in data.IBD_type] data.loc[:, 'ibd22'] = data.ibd21 data.loc[:, 'bucket_id'] = [mmh3.hash(id1) % denominator for id1 in data.id1] to_write = data.loc[:, ['bucket_id', 'id1', 'id2', 'chrom', 'phys_pos', 'zero1', 'marker_count', 'genetic_seg_length', 'cM', 'zero2', 'ibd21', 'ibd22']] for bucket_id, group in to_write.groupby('bucket_id'): dest_file = os.path.join(dest_dir, f'{bucket_id}.tsv.gz') group.drop('bucket_id', inplace=True, axis='columns') group.to_csv(dest_file, index=False, header=None, sep='\t', mode='a', compression='gzip') def split_by_id(input_ibd: str, samples_count: int, dest_dir: str): names = [ 'sample1', 'sample2', 'chrom', 'phys_start_pos', 'phys_end_pos', 'IBD_type', 'genetic_start_pos', 'genetic_end_pos', 'genetic_seg_length', 'marker_count', 'error_count', 'error_density' ] read_chunksize = int(1e+6) samples_chunksize = 2000 denominator = samples_count // samples_chunksize + 1 for i, chunk in enumerate(pandas.read_csv(input_ibd, header=None, names=names, sep='\t', chunksize=read_chunksize)): if not chunk.empty: process_chunk_with_hash(chunk, denominator, dest_dir) logging.info(f'Chunk {i} of size {read_chunksize} was written to {dest_dir} and split into {denominator} buckets') else: logging.info(f'No IBD segments were found in {input_ibd}. Nothing was written to {dest_dir}') if __name__ == '__main__': try: snakemake except NameError: Snakemake = namedtuple('Snakemake', ['input', 'output', 'log']) snakemake = Snakemake( input={ 'ibd': '/media/data1/relatives/runs/run100k/ibis/merged_ibis.seg', 'fam': '/media/data1/relatives/runs/run100k/preprocessed/data.fam'}, output={'bucket_dir': '/media/data1/relatives/test100koutput.txt'}, log=['/media/data1/relatives/test100koutput.log'] ) ibd = snakemake.input['ibd'] bucket_dir = snakemake.output['bucket_dir'] if not os.path.exists(bucket_dir): os.makedirs(bucket_dir) for file in os.listdir(bucket_dir): os.remove(os.path.join(bucket_dir, file)) samples = pandas.read_table(snakemake.input['fam']) samples_count = samples.shape[0] + 1 # there is no header in the file print(snakemake.log) logging.basicConfig(filename=snakemake.log[0], level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') split_by_id(ibd, samples_count, bucket_dir) |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 | import pandas import logging import mmh3 from typing import List import os from collections import namedtuple def process_chunk_with_hash(data: pandas.DataFrame, denominator: int, dest_dir: str): logging.info(f'DTYPES of rapid frame are {data.dtypes}') data.loc[:, 'id1'] = data.sample1.str.replace(':', '_') + ' ' + data.sample1.str.replace(':', '_') data.loc[:, 'id2'] = data.sample2.str.replace(':', '_') + ' ' + data.sample2.str.replace(':', '_') data.loc[:, 'phys_pos'] = [f'{start} {end}' for start, end in zip(data.genetic_start_pos, data.genetic_end_pos)] data.loc[:, 'zero1'] = '0 0' data.loc[:, 'zero2'] = '0' data.loc[:, 'cM'] = 'cM' data.loc[:, 'ibd21'] = 0 data.loc[:, 'ibd22'] = 0 data.loc[:, 'bucket_id'] = [mmh3.hash(id1) % denominator for id1 in data.id1] data.loc[:, 'marker_count'] = data.ending_site - data.starting_site to_write = data.loc[:, ['bucket_id', 'id1', 'id2', 'chrom', 'phys_pos', 'zero1', 'marker_count', 'genetic_seg_length', 'cM', 'zero2', 'ibd21', 'ibd22']] for bucket_id, group in to_write.groupby('bucket_id'): dest_file = os.path.join(dest_dir, f'{bucket_id}.tsv') group.drop('bucket_id', inplace=True, axis='columns') group.to_csv(dest_file, index=False, header=None, sep='\t', mode='a') def split_by_id(input_ibd: str, samples_count: int, dest_dir: str): """ <chr_name> <sample_id1> <sample_id2> <hap_id1> <hap_id2> <starting_pos_genomic> <ending_pos_genomic> <genetic_length> <starting_site> <ending_site> """ names = [ 'chrom', 'sample1', 'sample2', 'hap_id1', 'hap_id2', 'genetic_start_pos', 'genetic_end_pos', 'genetic_seg_length', 'starting_site', 'ending_site' ] read_chunksize = int(1e+6) samples_chunksize = 1000 denominator = samples_count // samples_chunksize + 1 for i, chunk in enumerate(pandas.read_csv(input_ibd, header=None, names=names, sep='\t', chunksize=read_chunksize)): if not chunk.empty: process_chunk_with_hash(chunk, denominator, dest_dir) logging.info(f'Chunk {i} of size {read_chunksize} was written to {dest_dir} and split into {denominator} buckets') else: logging.info(f'No IBD segments were found in {input_ibd}. Nothing was written to {dest_dir}') if __name__ == '__main__': try: snakemake except NameError: Snakemake = namedtuple('Snakemake', ['input', 'output', 'log']) snakemake = Snakemake( input={ 'ibd': '/media/data1/relatives/runs/run100k/ibis/merged_ibis.seg', 'fam': '/media/data1/relatives/runs/run100k/preprocessed/data.fam'}, output={'bucket_dir': '/media/data1/relatives/test100koutput.txt'}, log=['/media/data1/relatives/test100koutput.log'] ) ibd = snakemake.input['ibd'] bucket_dir = snakemake.output['bucket_dir'] if not os.path.exists(bucket_dir): os.makedirs(bucket_dir) for file in os.listdir(bucket_dir): os.remove(os.path.join(bucket_dir, file)) samples = pandas.read_table(snakemake.input['fam']) samples_count = samples.shape[0] + 1 # there is no header in the file print(snakemake.log) logging.basicConfig(filename=snakemake.log[0], level=logging.DEBUG, format='%(levelname)s:%(asctime)s %(message)s') split_by_id(ibd, samples_count, bucket_dir) |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 | import allel import zarr import numpy import os import time def write_ped_buffer(buffer: dict, out: str): start = time.time() with open(out, 'a') as ped: lines = [] for sample, gt in buffer.items(): #gt_str = numpy.array2string(gt, separator=' ', max_line_width=gt.shape[0], threshold=gt.shape[0]) line = ' '.join(['0', sample, '0', '0', '0', '0']) + ' ' + ' '.join(gt) + '\n' print(sample, len(gt)) lines.append(line) ped.writelines(lines) end = time.time() print(f'write_ped_buffer took {end - start} seconds') def write_map_file(ids, chrom, positions, out: str): with open(out, 'w') as file: for _id, _chr, pos in zip(ids, chrom, positions): file.write('\t'.join([_chr, _id, '0', str(pos)]) + '\n') if __name__ == '__main__': start = time.time() vcf = snakemake.input['vcf'][0] zarr_cache = snakemake.params['zarr'] ped = snakemake.output['ped'] map_file = snakemake.output['map_file'] ''' vcf = 'test_data/imputed_chr9.vcf.gz' zarr_cache = 'test_data/imputed_chr9.zarr' ped = 'test_data/imputed_chr9.ped' map_file = 'test_data/imputed_chr9.map' ''' ''' vcf = '/home/ag3r/vcf_to_ped/testsample.vcf.gz' zarr_cache = '/home/ag3r/vcf_to_ped/testsample.zarr' ped = '/home/ag3r/vcf_to_ped/testsample.ped' map_file = '/home/ag3r/vcf_to_ped/testsample.map' ''' if os.path.exists(ped): os.remove(ped) allel.vcf_to_zarr( input=vcf, output=zarr_cache, overwrite=True, fields=['variants/CHROM', 'variants/ID', 'variants/POS', 'variants/REF', 'variants/ALT', 'calldata/GT', 'samples'] ) array = zarr.open_group(zarr_cache, mode='r') samples_in_chunk = 100 ids = array['variants/ID'][:] chrom = array['variants/CHROM'][:] positions = array['variants/POS'][:] write_map_file(ids, chrom, positions, map_file) # only for the single chromosome! print(f'written total of {len(positions)} variants from chr{chrom[0]} to {map_file}') for sample_start in range(0, len(array['samples']), samples_in_chunk): sample_end = min(len(array['samples']), sample_start + samples_in_chunk) samples = array['samples'][sample_start: sample_end] ped_buffer = {} # [variants] ref = array['variants']['REF'][:].reshape(-1, 1) # [variants, alt_count] alt = array['variants']['ALT'][:] # [variants, alt_count + 1] alleles = numpy.hstack([ref, alt]) alleles[alleles == ''] = ' ' # [variants, samples, 2] # 2 because we use biallelic only gt = array['calldata']['GT'][:, sample_start: sample_end, :] for sample_idx, sample in enumerate(samples): # [variants, 2] line = numpy.zeros((gt.shape[0], 2), dtype=alleles.dtype) # [variants] contains 0 or 1, i.e. count of reference alleles in corresponding haplotype (fwd or bck) fwd_idx = gt[:, sample_idx, 0].astype(bool) bck_idx = gt[:, sample_idx, 1].astype(bool) #print(gt.shape, alleles.shape, gt[:, sample_idx, 0].shape) # [variants] ref_line contains ref variants, alt_line contains alf_variants ref_line, alt_line = alleles[~fwd_idx, 0], alleles[fwd_idx, 1] # we are setting first haplotype ref and alt variants in corresponding positions line[~fwd_idx, 0] = ref_line line[fwd_idx, 0] = alt_line ref_line, alt_line = alleles[~bck_idx, 0], alleles[bck_idx, 1] line[~bck_idx, 1] = ref_line line[bck_idx, 1] = alt_line line_array = line.ravel() #print(line_array[:5], line_array.shape) #for variant_idx in range(gt.shape[0]): # fwd_idx, bck_idx = gt[variant_idx, sample_idx, 0], gt[variant_idx, sample_idx, 1] # fwd = alleles[variant_idx, fwd_idx] # ped_line.append(fwd) # bck = alleles[variant_idx, bck_idx] # ped_line.append(bck) ped_buffer[sample] = line_array write_ped_buffer(ped_buffer, ped) print(f'written {len(ped_buffer)} samples to {ped}') end = time.time() print(f'total execution time is {end - start:.4f} seconds') print(f'written total of {len(array["samples"])} and {len(array["variants"]["REF"])} variants to {ped}') |
Support
- Future updates
Related Workflows





