bkohrn / Duplex-Seq-Pipeline

A standalone end-to-end data analysis pipeline for Duplex Sequencing

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Duplex Sequencing Pipeline

Brendan Kohrn, April 26, 2021
Duplex Sequencing is copyright Scott Kennedy and Larwrence Loeb, University of Washington, Seattle, WA, USA

Table of Contents:

  1. Glossary
  2. Dependencies
  3. Setup
  4. Genome Setup
  5. Contaminant database setup
  6. Bed and Interval_list file preparation
  7. Configuration File creation
  8. Recovery script creation
  9. Running pipelines
  10. Output file descriptions
  11. Extra BAM tags:
  12. Testing the pipeline
  13. Full and partial reruns
  14. Unlocking following a power failure

1: Glossary

Single Stranded Consensus Sequence (SSCS)

A construct created by comparing multiple reads and deciding ambiguities by simple majority.

Duplex Consensus Sequence (DCS)

A construct created by comparing two SSCSs.

Family

A group of reads that shares the same tag sequence.

2: Dependencies:

This pipeline is known to work with the following versions of the following required programs:

  • Python3.6+
  • Snakemake>=5.25.0
  • Pandas
  • Miniconda=4.7.*
  • bwa=0.7.17.* (for genome setup)
  • ncbi-blast=2.9.0 (installed separately, for contaminant database setup; note that while an older version can be used for database setup, a newer version CANNOT be used for database setup.)
  • mamba>=0.5.1 (needed for environment setup; install through conda)
  • wget (on macOS, install using homebrew; present by default on linux)

Once Python3.6 is installed, snakemake and pandas can be installed using pip3 or using whatever package manager you're using. NOTE: For MacOS X, use the latest Snakemake. Blast can be downloaded in any of several ways, including some package managers (Ubuntu: sudo apt-get install ncbi-blast+). It can also be installed using conda if desired.

The recommended install method and order is:

  1. Install Miniconda
  2. Use conda to install Python3 and Mamba
  3. Use Mamba to install Snakemake, bwa, blast, and pandas, invoking both the bioconda and conda-forge channels.

3: Setup:

Find the location where you want the pipeline to be located and clone the pipeline using git:

git clone https://github.com/KennedyLabUW/Duplex-Seq-Pipeline.git

change into the directory, and run:

bash setupDS.sh MAX_CORES

where MAX_CORES is the maximum number of cores you want the pipeline to be able to use. This will create a test configuration file in the test directory (test/testConfig.csv), a basic configuration file (DS_progConfig.yaml), a run script (DS), a DAG-creation script (DS-dag), a reset script (DS-clean), and an unlock script (DS-unlock). After this, with the exception of setting up the Genomes (See Section 4) and the (optional) blast contamination database (See Section 5), you should be able to run the Duplex-Seq pipeline using

bash /path/to/Duplex-Seq-Pipeline/DS CONFIG_CSV.csv

where CONFIG_CSV.csv is a configuration CSV file generated as described below.

If you want, you can also add the DS and DS-clean scripts to your path; this will simplify invocation if you only have a single release version on your system.

4: Genome setup

Put genomes in an easily findable location, such as our references directory (i.e. ~/bioinformatics/reference).

Many genomes can be downloaded from UCSC (http://hgdownload.soe.ucsc.edu/downloads.html). In order to download genomes from there, you will need to download the twoBitToFa program from the appropriate Utilities directory. twoBitToFa has the following syntax (Copied from UCSC):

twoBitToFa - Convert all or part of .2bit file to fasta
usage:
   twoBitToFa input.2bit output.fa
options:
   -seq=name       Restrict this to just one sequence.
   -start=X        Start at given position in sequence (zero-based).
   -end=X          End at given position in sequence (non-inclusive).
   -seqList=file   File containing list of the desired sequence names 
                   in the format seqSpec[:start-end], e.g. chr1 or chr1:0-189
                   where coordinates are half-open zero-based, i.e. [start,end).
   -noMask         Convert sequence to all upper case.
   -bpt=index.bpt  Use bpt index instead of built-in one.
   -bed=input.bed  Grab sequences specified by input.bed. Will exclude introns.
   -bedPos         With -bed, use chrom:start-end as the fasta ID in output.fa.
   -udcDir=/dir/to/cache  Place to put cache for remote bigBed/bigWigs.

Sequence and range may also be specified as part of the input file name using the syntax:

      /path/input.2bit:name
   or
      /path/input.2bit:name
   or
      /path/input.2bit:name:start-end

Once you have downloaded a genome and converted it into FASTA format, it needs to be indexed. To do this, open a terminal and navigate to your the directory containing your genome. Note that in the following commands, the word "genome.fasta" should be replaced with the file name of your genome. These commands must be run in the same directory as the reference genome

    bwa index genome.fasta  
    samtools faidx genome.fasta  
    # /Path/To/PicardTools should be the path to wherever you put your picard tools jar file.    
    # Also, genome.dict should match the name of your fasta (e.g. hg19.fasta to hg19.dict)  
    java -jar /Path/To/PicardTools/picard.jar CreateSequenceDictionary R=genome.fasta O=genome.dict  

At the moment, this pipeline does not support compressed genomes.

5: Contaminant Database Setup:

The Duplex-Seq pipeline is designed to use a local NCBI Blast instance to detect and remove potential contamination from non-target species and identify issues arising from pseudogenes. This step is optional, but requires a valid NCBI Blast database file name. To run without BLAST, enter "NONE" (with any capitalization) in the blast_db field in the config.csv file.

To construct your contaminant database, if desired, first decide on a list of species you want to monitor for contaminants. A suggested starting list is:

  • Human (hg38)
  • Mouse (mm10)
  • Rat (rn6)
  • C. elegens (ce11)
  • Yeast (sacCer3)
  • Fruit Fly (dm6)
  • Cow (bosTau9)
  • Dog (canFam3)

It is important that any genome you plan to use for alignment be included in this database, in the same version (e.g. if your alignment genome uses UCSC chromosome names, your genome in the database cannot use NCBI chromosome names, but must also use UCSC chromosome names). It should also be noted that, while older versions of BLAST can be used for database setup, setup should use at the latest version 2.9.0; databases constructed with later BLAST versions may not work with the pipeline.

Database setup consists of two steps:

  1. Sub-database Creation:

Create the database using:

makeblastdb \  
-dbtype nucl \  
-title GENOME \  
-out GENOME_db \  
-in GENOME.fa \  
-taxid TAXID  

where GENOME.fa is the input fasta file with the genome, and TAXID is the NCBI taxonomy ID for the species associated with the genome. TaxIDs can be found using the NCBI taxonomy website (https://www.ncbi.nlm.nih.gov/taxonomy).

After this, create a .nal file using:

blastdb_aliastool -dblist "GENOME_db" \  
-dbtype nucl -out GENOME_db -title "GENOME"

The blastDbSetup.sh script can be used to automate these steps. It can be run with:

bash /path/to/pipeline/setupBlastDb/blastDbSetup.sh GENOME.fa TAXON_ID
  1. Full database creation:

Once you've created all your sub-databases (e.g. GENOME1, GENOME2, GENOME3, ..., GENOME_N), create a .nal file to represent the full database using:

blastdb_aliastool -dblist "GENOME1_db GENOME2_db GENOME3_db ... GENOME_N_db" \  
-dbtype nucl -out contaminant_db -title "Contaminant Database"

If, at a later time, you need to change your contaminant database, you then only need to rebuild the modified portion, and then update this file to reflect that.

6: Bed file preparation

A bed file is a file which details regions of the genome in which you are interested. The syntax for bed files is described at https://genome.ucsc.edu/FAQ/FAQformat.html#format1.

If you know which genes you are targeting and are using a common published genome, the target bed file can be downloaded from the UCSC Table Browser (https://genome.ucsc.edu/cgi-bin/hgTables). Otherwise, it can be created using results from a BLAST search or any other method you like. This pipeline does not currently support overlapping intervals, but does support blocks as described in the bed spec.

This pipeline can also optionally use a masking bed file, for if there are genomic regions where you know there will be a high number of artifactual variants, or which you would like to ignore for other reasons. The masking bed file only needs to contain the first three bed columns (chrom, start, and stop), and will ignore any other columns provided.

7: Configuration file creation:

Use the ConfigTemplate to create a new file with the appropriate headers. For each row, fill in the information about a particular sample. Note that in most cases, "path" refers to an absolute path, with the exception of "in1" and "in2", where you just need the name of the input file. The "blast_db" option is the path to the appropriate .nal or .nhr file, minus the .nal or .nhr extension, but can also be specified as "none" to skip BLAST filtering.

Header Required or Default Information
sample Required A unique identifier for a sample; this will be used to name all output files for this sample
rglb Required Read Group Library Identifier
rgpl Required Read Group Platform; usually Illumina
rgpu Required Read Group Platform Unit
rgsm Required Read Group Sample
reference Required The path to the prepared reference genome to use with this sample.
target_bed Required A bed file showing where the targets are for this particular sample
maskBed NONE A bed file to use for masking variants.
blast_db NONE The blast database to use for contaminant filtering; must include your target genome, if used.
targetTaxonId 9606 The taxon ID of the species you are expecting to be present in the sample.
baseDir Required The directory the input files are in, and where the output files will be created.
in1 Required The read1 fastq (or fastq.gz, or fq.gz, or fq) file for this sample. Note that this is just the name of the file, and not the full path.
in2 Required The read2 fastq (or fastq.gz, or fq.gz, or fq) file for this sample. Note that this is just the name of the file, and not the full path.
mqFilt 0 A threshold for mapping quality filtering, if desired.
minMem 0 The minimum number of reads that must be in a family for consensus making
maxMem 200 The maximum number of reads in a family the consensus maker should consider.
cutOff 0.9 The threshold for consensus making; the consensus maker will require at least this much agreement on a per base pair level.
nCutOff 1 The maximum proportion of N bases in an output consensus sequence.
umiLen 8 The length of the UMI in this sample
spacerLen 1 The length of the spacer sequence in this sample
locLen 10 The localization length to use for this sample
readLen 101 The length of a read for this sample
adapterSeq "ANNNNNNNNAGATCGGAAGAG" The adapter sequence used in library preperation, with UMI bases as Ns, and spacer sequence included. Alternatively, a fasta file with all possible UMI-adapter conbinations. Used by cutadapt for adapter clipping.
clipBegin 7 How many bases to clip off the 5' end of the read
clipEnd 0 How many bases to clip off the 3' end of the read
minClonal 0 The minimum clonality to use for count_muts generation
maxClonal 0.1 The maximum clonality to use for count_muts generation
minDepth 100 The minimum depth to use for count_muts generation
maxNs 1 The maximum proportion of N bases to use for count_muts generation
recovery "noRecovery.sh" The recovery script to use in attempting to recover ambiguously mapped reads (as determine by blast alignment vs bwa alignment). Recovery script creation is discussed in 8; below.
cluster_dist 10 How close together variants have to be to be considered 'clustered'
cm_outputs "GB" Select which sections of the countmuts to output, in addition to 'OVERALL'. String of one or more of 'G', 'B', and 'N'. G -> output GENE sections for each bed line; B -> output 'BLOCK' sections for each block in the bed line (if present); 'N' -> Only output overall frequencies. Overrides all other options.
cm_sumTypes "GT" How to calculate OVERALL and GENE blocks for countmuts output. The first character controls summing for overall: G -> OVERALL = sum(GENEs); B -> OVERALL = sum(BLOCKs). In sum(GENEs) mode, this will ignore BLOCKs for the purposes of calculating OVERALL. The second character controls summing for each GENE: T -> GENE = Whole gene, ignoring BLOCKs; B -> GENE = sum(BLOCKs).
cm_filters "none" Select which filters to apply during frequency calculation. These filters will also be applied during muts_per_cycle calculation. Available filters are: SNP, near_indel, clustered, masked, low_depth.
runSSCS false true or false; whether to do full analysis for SSCS data.
rerun_type 0 (Required for rerun) What type of rerun you want to do. 0 -> no rerun; 1 -> rerun variant caller; 2 -> rerun postBlastRecovery; 3 -> rerun BLAST and alignment; 4 -> rerun consensus maker.

Save the file as a .csv file with unix line endings (LF).

8: Recovery script creation

As part of its operation, this pipeline filters out correct-species reads where the blast mapping and bwa mapping positions disagree or where blast is unable to determine conclusively where the read maps (i.e.E-scores are the same). There is a step which provides an option to recover those reads by using a user-generated bash script. Currently, we use a bash script to call a python script which will actually accomplish the recovery, but this functionality may be changed in the future. In general, these scripts must:

  1. accept ambiguous reads as $1
  2. accept non-ambiguous reads as $2
  3. accept reads labeled as incorrect species as $3
  4. take a name base for output files as $4
  5. take a basePath for location of script files as $5

The script must create the following output files:

  • ${4}.recovered.temp.bam
  • ${4}.ambig.bam
  • ${4}.wrongSpecies.bam

All script files must be stored in scripts/RecoveryScripts.

Any packages required by your recovery script can be added to the envs/DS_env_recovery.yaml file. By default, this contains the conda packages (from bioconda or conda-forge):

  • Python 3.6
  • samtools
  • pysam
  • snakemake
  • pandas
  • regex
  • biopython
  • pandoc

Note that if you decide to write your own recovery scripts, you are responsible for ensuring that your recovery scripts actually work as intended, and do not break your computer. We are unable to guarantee that any sample run with a recovery script other than one of the ones provided with the pipeline will produce accurate data.

9: Running pipelines:

The pipeline can be run using the DS command created by the setup script:

/path/to/Duplex-Seq-Pipeline/DS CONFIG_CSV.csv

10: Output file descriptions:

The pipeline will create a set of summary files covering all samples, as well as a file directory structure for each sample. The summary files are:

File Name Description
summary.csv A csv file with summary metrics for all samples.
summaryDepth.pdf A pdf file containing depth per target plots for all samples.
summaryFamilySize.pdf A pdf file containing family size plots for all samples.
summaryInsertSize.pdf A pdf file containing insert size plots for all samples.
summaryMutsByCycle.pdf A pdf file containing non-SNP mutations per cycle for all samples.

This directory structure after a run looks like this:

.
└──SAMP_DIR
    ├── Final
    │   ├── dcs
    │   │   ├── FilteredReads
    │   │   │   ├── SAMPLE_dcs.postRecovery.ambig.bam
    │   │   │   ├── SAMPLE_dcs.postRecovery.ambig.bam.bai
    │   │   │   ├── SAMPLE_dcs.postRecovery.wrongSpecies.bam
    │   │   │   └── SAMPLE_dcs.postRecovery.wrongSpecies.bam.bai
    │   │   ├── SAMPLE.dcs.countmuts.csv
    │   │   ├── SAMPLE.dcs.final.bam
    │   │   ├── SAMPLE.dcs.final.bam.bai
    │   │   ├── SAMPLE.dcs.mutated.bam
    │   │   ├── SAMPLE.dcs.mutated.bam.bai
    │   │   ├── SAMPLE.dcs.snps.vcf
    │   │   └── SAMPLE.dcs.vcf
    │   ├── SAMPLE.report.html
    │   └── sscs
    │       ├── SAMPLE.sscs.countmuts.csv
    │       ├── SAMPLE.sscs.final.bam
    │       ├── SAMPLE.sscs.final.bam.bai
    │       ├── SAMPLE.sscs.mutated.bam
    │       ├── SAMPLE.sscs.mutated.bam.bai
    │       ├── SAMPLE.sscs.snps.vcf
    │       └── SAMPLE.sscs.vcf
    ├── Intermediate
    │   ├── ConsensusMakerOutputs
    │   │   ├── SAMPLE_aln_seq1.fq.gz
    │   │   ├── SAMPLE_aln_seq2.fq.gz
    │   │   ├── SAMPLE_read1_dcs.fq.gz
    │   │   ├── SAMPLE_read1_sscs.fq.gz
    │   │   ├── SAMPLE_read2_dcs.fq.gz
    │   │   └── SAMPLE_read2_sscs.fq.gz
    │   ├── preVariantCalling
    │   │   ├── SAMPLE.dcs.prevar.bam
    │   │   └── SAMPLE.sscs.prevar.bam
    │   └── postBlast
    │       ├── FilteredReads
    │       │   ├── SAMPLE_dcs.ambig.sort.bam
    │       │   ├── SAMPLE_dcs.ambig.sort.bam.bai
    │       │   ├── SAMPLE_dcs.wrongSpecies.sort.bam
    │       │   └── SAMPLE_dcs.wrongSpecies.sort.bam.bai
    │       ├── SAMPLE_dcs.blast.xml
    │       ├── SAMPLE_dcs.preBlast.mutated.bam
    │       └── SAMPLE_dcs.preBlast.unmutated.bam
    ├── logs
    │   └── Log files
    ├── testSeq1.fastq.gz
    ├── testSeq2.fastq.gz
    └── Stats
        ├── data
        │   ├── SAMPLE_cmStats.txt
        │   ├── SAMPLE.dcs_ambiguity_counts.txt
        │   ├── SAMPLE.dcs.iSize_Metrics.txt
        │   ├── SAMPLE.dcs_MutsPerCycle.dat.csv
        │   ├── SAMPLE.dcs.mutsPerRead.txt
        │   ├── SAMPLE.sscs_MutsPerCycle.dat.csv
        │   ├── SAMPLE.sscs.mutsPerRead.txt
        │   ├── SAMPLE.dcs.depth.txt
        │   ├── SAMPLE.dcs.depth.summary.csv
        │   ├── SAMPLE.sscs.depth.txt
        │   ├── SAMPLE.sscs.depth.summary.csv
        │   ├── SAMPLE_dcs.speciesComp.txt
        │   ├── SAMPLE_mem.dcs.sort.flagstats.txt
        │   ├── SAMPLE_mem.sscs.sort.flagstats.txt
        │   ├── SAMPLE.dcs.endClip.metrics.txt
        │   ├── SAMPLE.dcs.overlapClip.metrics.txt
        │   ├── SAMPLE.sscs.endClip.metrics.txt
        │   ├── SAMPLE.sscs.overlapClip.metrics.txt
        │   ├── SAMPLE_onTargetCount.txt
        │   ├── SAMPLE.sscs_onTargetCount.txt
        │   ├── SAMPLE.dcs_onTargetCount.txt
        │   ├── SAMPLE.tagstats.txt
        │   └── SAMPLE.temp.sort.flagstats.txt
        ├── SAMPLE.report.ipynb
        └── plots
            ├── SAMPLE.dcs.iSize_Histogram.png
            ├── SAMPLE.dcs.mutsPerRead.png
            ├── SAMPLE.dcs.targetCoverage.png
            ├── SAMPLE.dcs_BasePerPosInclNs.png
            ├── SAMPLE.dcs_BasePerPosWithoutNs.png
            ├── SAMPLE.sscs.mutsPerRead.png
            ├── SAMPLE.sscs_BasePerPosInclNs.png
            ├── SAMPLE.sscs_BasePerPosWithoutNs.png
            ├── SAMPLE_fam_size_relation.png
            └── SAMPLE_family_size.png

File descriptions are as follows:

Directory File name Description When Generated
SAMP_DIR SAMPLE_seq1.fastq.gz Input read 1 file Input
SAMP_DIR SAMPLE_seq2.fastq.gz Input read 2 file Input
SAMP_DIR Final Directory containing final bam and vcf files Always
SAMP_DIR/Final dcs Directory containing final dcs files Always
SAMP_DIR/Final/dcs FilteredReads Directory containing reads still filtered after postBlastRecovery blast_db!=NONE
SAMP_DIR/Final/dcs/FilteredReads SAMPLE.dcs.postRecovery.ambig.bam File containing reads still considered ambiguous after postBlastRecovery. May be empty. blast_db!=NONE
SAMP_DIR/Final/dcs/FilteredReads SAMPLE.dcs.postRecovery.ambig.bam.bai Index for SAMPLE.dcs.postRecovery.ambig.bam blast_db!=NONE
SAMP_DIR/Final/dcs/FilteredReads SAMPLE.dcs.postRecovery.wrongSpecies.bam File containing reads still considered ambiguous after postBlastRecovery. May be empty. blast_db!=NONE
SAMP_DIR/Final/dcs/FilteredReads SAMPLE.dcs.postRecovery.wrongSpecies.bam.bai Index for SAMPLE.dcs.postRecovery.wrongSpecies.bam blast_db!=NONE
SAMP_DIR/Final/dcs SAMPLE.dcs.countmuts.csv Countmuts file, showing a summary of mutation data for DCS reads. Always
SAMP_DIR/Final/dcs SAMPLE.dcs.final.bam Final file for DCS reads, including all reads that overlap the bed file. Always
SAMP_DIR/Final/dcs SAMPLE.dcs.final.bam.bai Index for final DCS reads. Always
SAMP_DIR/Final/dcs SAMPLE.dcs.mutated.bam File containing DCS reads with non-SNP mutations Always
SAMP_DIR/Final/dcs SAMPLE.dcs.mutated.bam.bai Index for DCS mutated reads file. Always
SAMP_DIR/Final/dcs SAMPLE.dcs.snps.vcf VCF file containing SNPs overlapping bed file in DCS. Always
SAMP_DIR/Final/dcs SAMPLE.dcs.vcf VCF file containing all variants overlapping bed file in DCS. Always
SAMP_DIR/Final SAMPLE.report.html Summary report for this sample Always
SAMP_DIR/Final sscs Directory containing final SSCS files Always
SAMP_DIR/Final/sscs SAMPLE.sscs.countmuts.csv Countmuts file, showing a summary of mutation data for SSCS reads. runSscs=True
SAMP_DIR/Final/sscs SAMPLE.sscs.final.bam Final file for SSCS reads, including all reads that overlap the bed file. Always
SAMP_DIR/Final/sscs SAMPLE.sscs.final.bam.bai Index for final SSCS reads. Always
SAMP_DIR/Final/sscs SAMPLE.sscs.mutated.bam File containing SSCS reads with non-SNP mutations runSscs=True
SAMP_DIR/Final/sscs SAMPLE.sscs.mutated.bam.bai Index for SSCS mutated reads file. runSscs=True
SAMP_DIR/Final/sscs SAMPLE.sscs.snps.vcf VCF file containing SNPs overlapping bed file in SSCS. runSscs=True
SAMP_DIR/Final/sscs SAMPLE.sscs.vcf VCF file containing all variants overlapping bed file in SSCS. runSscs=True
SAMP_DIR Intermediate Directory containing intermediate checkpointing files Always
SAMP_DIR/Intermediate ConsensusMakerOutputs Directory for post-consensus maker checkpoint files Always
SAMP_DIR/Intermediate/ConsensusMakerOutputs SAMPLE_aln_seq1.fq.gz Read 1 file for raw on-target determination Always
SAMP_DIR/Intermediate/ConsensusMakerOutputs SAMPLE_aln_seq2.fq.gz Read 2 file for raw on-target determination Always
SAMP_DIR/Intermediate/ConsensusMakerOutputs SAMPLE_read1_dcs.fq.gz DCS Read 1 File Always
SAMP_DIR/Intermediate/ConsensusMakerOutputs SAMPLE_read1_sscs.fq.gz SSCS Read 1 file Always
SAMP_DIR/Intermediate/ConsensusMakerOutputs SAMPLE_read2_dcs.fq.gz DCS Read 2 File Always
SAMP_DIR/Intermediate/ConsensusMakerOutputs SAMPLE_read2_sscs.fq.gz SSCS Read 2 file Always
SAMP_DIR/Intermediate preVariantCalling Directory for pre-variant calling checkpoint files Always
SAMP_DIR/Intermediate/preVariantCalling SAMPLE.dcs.prevar.bam DCS pre-variant file Always
SAMP_DIR/Intermediate/preVariantCalling SAMPLE.sscs.prevar.bam SSCS pre-variant file Always
SAMP_DIR/Intermediate/ postBlast Directory for post-BLAST checkpoint files. Only affects DCS. Always
SAMP_DIR/Intermediate/postBlast FilteredReads Directory for reads that got filtered out of DCS processing due to BLAST analysis indicating that they were either the wrong species or ambiguously mapped. Always
SAMP_DIR/Intermediate/postBlast/FilteredReads SAMPLE_dcs.ambig.sort.bam Reads that were filtered out due to ambiguous mapping according to BLAST alignment. Always
SAMP_DIR/Intermediate/postBlast/FilteredReads SAMPLE_dcs.ambig.sort.bam.bai Index for ambiguous reads file Always
SAMP_DIR/Intermediate/postBlast/FilteredReads SAMPLE_dcs.wrongSpecies.sort.bam Reads that were filtered out due to BLAST alignment indicating that they were from the wrong species, or where species of origin could not be determined. Always
SAMP_DIR/Intermediate/postBlast/FilteredReads SAMPLE_dcs.wrongSpecies.sort.bam.bai Index for wrong-species file Always
SAMP_DIR/Intermediate/postBlast SAMPLE_dcs.blast.xml BLAST xml output Always
SAMP_DIR/Intermediate/postBlast SAMPLE_dcs.preBlast.mutated.bam DCS with potential non-SNP variants that were submitted to BLAST. Always
SAMP_DIR/Intermediate/postBlast SAMPLE_dcs.preBlast.unmutated.bam DCS reads without non-SNP variants. Always
SAMP_DIR logs Directory containing log files for this sample. Log files are currently empty, but will be filled later. Always
SAMP_DIR Stats Directory containing statistics files Always
SAMP_DIR/Stats data Directory containing statistics data files. Always
SAMP_DIR/Stats/data SAMPLE_cmStats.txt Statistics from the Consensus Maker Always
SAMP_DIR/Stats/data SAMPLE.dcs_ambiguity_counts.txt Statistics on ambiguity counts Always
SAMP_DIR/Stats/data SAMPLE.dcs.iSize_Metrics.txt Statistics on insert size in DCS Always
SAMP_DIR/Stats/data SAMPLE.dcs_MutsPerCycle.dat.csv Statistics file for non-SNP mutations per cycle in DCS reads Always
SAMP_DIR/Stats/data SAMPLE.dcs.mutsPerRead.txt Statistics file for non-SNP mutations per read in DCS reads Always
SAMP_DIR/Stats/data SAMPLE.sscs_MutsPerCycle.dat.csv Text data of error rate per cycle in unclipped SSCS runSscs=True
SAMP_DIR/Stats/data SAMPLE.sscs.mutsPerRead.txt Statistics file for non-SNP mutations per read in SSCS reads runSscs=True
SAMP_DIR/Stats/data SAMPLE.dcs.depth.txt Per-base coverage and N counts for final DCS Always
SAMP_DIR/Stats/data SAMPLE.dcs.depth.summary.csv Per-bed region min, mean, median, and max non-zero depth for final DCS. Always
SAMP_DIR/Stats/data SAMPLE.sscs.depth.txt Per-base coverage and N counts for final SSCS runSscs=True
SAMP_DIR/Stats/data SAMPLE.sscs.depth.summary.csv Per-bed region min, mean, median, and max non-zero depth for final SSCS. runSscs=True
SAMP_DIR/Stats/data SAMPLE_dcs.speciesComp.txt File containing species assignment data for DCS reads Always
SAMP_DIR/Stats/data SAMPLE_mem.dcs.sort.flagstats.txt Initial alignment statistics for DCS reads Always
SAMP_DIR/Stats/data SAMPLE_mem.sscs.sort.flagstats.txt Initial alignment statistics for SSCS reads Always
SAMP_DIR/Stats/data SAMPLE.dcs.endClip.metrics.txt Statistics on fixed end clipping in DCS Always
SAMP_DIR/Stats/data SAMPLE.dcs.overlapClip.metrics.txt Statistics on overlap clipping in DCS Always
SAMP_DIR/Stats/data SAMPLE.sscs.endClip.metrics.txt Statistics on fixed end clipping in SSCS runSscs=True
SAMP_DIR/Stats/data SAMPLE.sscs.overlapClip.metrics.txt Statistics on overlap clipping in SSCS runSscs=True
SAMP_DIR/Stats/data SAMPLE_onTargetCount.txt Raw on target statistics Always
SAMP_DIR/Stats/data SAMPLE.sscs_onTargetCount.txt SSCS on target statistics Always
SAMP_DIR/Stats/data SAMPLE.dcs_onTargetCount.txt DCS on target statistics Always
SAMP_DIR/Stats/data SAMPLE.tagstats.txt Family size data (in text form) Always
SAMP_DIR/Stats/data SAMPLE.temp.sort.flagstats.txt Statistics on initial read counts Always
SAMP_DIR/Stats SAMPLE.report.ipynb iPython notebook for the HTML report Always
SAMP_DIR/Stats plots Directory containing statistics plots. Always
SAMP_DIR/Stats/plots SAMPLE.dcs.iSize_Histogram.png Histogram of insert size metrics for un-clipped DCS Always
SAMP_DIR/Stats/plots SAMPLE.dcs.mutsPerRead.png Plot of mutations per read in DCS Always
SAMP_DIR/Stats/plots SAMPLE.dcs.targetCoverage.png Plot of per-target coverage in DCS Always
SAMP_DIR/Stats/plots SAMPLE.dcs_BasePerPosInclNs.png Plot of error rate per cycle for DCS, including Ns Always
SAMP_DIR/Stats/plots SAMPLE.dcs_BasePerPosWithoutNs.png Plot of error rate per cycle for DCS Always
SAMP_DIR/Stats/plots SAMPLE.sscs.mutsPerRead.png Plot of mutations per read in SSCS runSscs=True
SAMP_DIR/Stats/plots SAMPLE.sscs_BasePerPosInclNs.png Plot of error rate per cycle SSCS, including Ns runSscs=True
SAMP_DIR/Stats/plots SAMPLE.sscs_BasePerPosWithoutNs.png Plot of error rate per cycle for SSCS runSscs=True
SAMP_DIR/Stats/plots SAMPLE_fam_size_relation.png Plot of relationship between a:b and b:a families Always
SAMP_DIR/Stats/plots SAMPLE_family_size.png Plot of family size distribution Always

11: Extra BAM tags:

Tag Type Meaning
XF String In SSCS, represents family size. In DCS, colon-deliniated list of family sizes for ab1:ba2 or ba1:ab2.
t0 Integer With BLAST, TaxID of the species the read matched most closely. Note that negative numbers have special meanings; see below.
t# Integer With BLAST, TaxID of the #th BLAST hit.
c# String With BLAST, the chromosome of the #th BLAST hit.
p# Integer With BLAST, the position of the #th BLAST hit.
l# Integer With BLAST, the length of the #th BLAST hit.
YB String With BLAST, True if a read was BLASTed.
am integer With BLAST, the ambiguity code for the read. See below for meanings.

Some negative values for tag t0 have special meanings:

t0 value meaning
-1 Between species BLAST tie.
-3 No BLAST results.
-4 Read was submitted to BLAST, but BLAST did not attempt alignment.

Ambiguity codes have manings:

| am value | meaning |
| 0 | BLAST has 1 best match, matches bwa position |
| 1 | BLAST has 1 best match, does not match bwa position |
| 2 | BLAST has 2+ best matches, correct species |
| 3 | BLAST has 2+ best matches, at least one incorrect species |
| 4 | BLAST did not attempt alignment, or no blast matches |
| 5 | Not BLASTed |

12: Testing the pipeline

The newly setup pipeline can be tested using provided data and files located in the 'test' directory. To test the pipeline, change into the 'test' directory and invoking at the command prompt:

DS testConfig.csv

A final set of output reports can be found in the testData/Final directory and be compared to the reports in the ExpectedReports directory located in the parent test directory.

13: Full and partial reruns

Sometimes it may be necessary to rerun all or part of the pipeline for various reasons. In order to facilitate this, we have provided a script (DS-clean) which will prepare samples to rerun based on the "rerun_type" column in the config file.

The following table lists some of the reasons you might want to rerun all or part of the pipeline, and how much of the pipeline you want to rerun in those cases.

Issue Amount to rerun rerun_type
  • Wrong target bed file used
  • Wrong clipping parameters used
From pre-variant calling 1
Wrong target taxon ID used From post-blast 2
  • Wrong contaminant db used
  • Wrong reference genome used
  • Wrong adapter sequence used
From post-Consensus Maker 3
Wrong consensus making parameters used From beginning 4

To finish preparing for and executing a rerun, run:

/path/to/Duplex-Seq-Pipeline/DS-clean CONFIG_CSV.csv
/path/to/Duplex-Seq-Pipeline/DS CONFIG_CSV.csv

14: Unlocking following a power failure

In the event that pipeline execution is interrupted (such as by a power failure), the directory can be unlocked in order to restart using the provided DS-unlock script:

/path/to/Duplex-Seq-Pipeline/DS-unlock CONFIG_CSV.csv

About

A standalone end-to-end data analysis pipeline for Duplex Sequencing

License:Other


Languages

Language:HTML 92.7%Language:Python 6.8%Language:R 0.3%Language:Shell 0.2%