//wrapper should now handle all throwing and catching..
protected override void ProgramExecution()
{
var optionsUsed = _appOptionParser.ParsingResult.OptionsUsed;
var doNotPassToSubprocess = new List <string>()
{
"outFolder", "numProcesses", "exePath", "intermediateDir", "multiprocess", "chromosomes"
};
var cmdLineList = MultiProcessHelpers.GetCommandLineWithoutIgnoredArguments(optionsUsed, doNotPassToSubprocess);
var refNameMapping = new Dictionary <string, int>();
using (var bamReader = new BamReader(_options.InputBam))
{
var chroms = bamReader.GetReferenceNames();
foreach (var referenceName in chroms)
{
if (_options.Chromosomes != null && !_options.Chromosomes.ToList().Contains(referenceName))
{
continue;
}
refNameMapping.Add(referenceName, bamReader.GetReferenceIndex(referenceName));
}
}
var taskManager = new CliTaskManager(_options.NumProcesses);
var geminiProcessor = new GeminiMultiProcessor(_options, new CliTaskCreator());
var samtoolsWrapper = new SamtoolsWrapper(_options.GeminiOptions.SamtoolsPath, _options.GeminiOptions.IsWeirdSamtools);
geminiProcessor.Execute(taskManager, refNameMapping, cmdLineList, samtoolsWrapper);
}
private static string ProcessChromosome(Dictionary <string, int> chromRefIds, string outMultiPath, List <string> taskDirectories,
string chrom, GeminiMultiOptions options)
{
// TODO either officially deprecate non-multiprocess-processing and remove this, or consolidate this with the Gemini calling code from Gemini/Program.cs
var outdir = Path.Combine(outMultiPath, chrom);
var refId = chromRefIds[chrom];
var intermediate = string.IsNullOrEmpty(options.GeminiSampleOptions.IntermediateDir)
? null
: Path.Combine(options.GeminiSampleOptions.IntermediateDir, chrom);
var geminiSampleOptions = new GeminiSampleOptions
{
InputBam = options.InputBam,
OutputFolder = outdir,
OutputBam = Path.Combine(outdir, "out.bam"),
IntermediateDir = intermediate,
RefId = refId
};
// Gemini defaults different than stitcher defaults
options.StitcherOptions.NifyUnstitchablePairs = false;
// Set stitcher pair-filter-level duplicate filtering if skip and remove dups, to save time
options.StitcherOptions.FilterDuplicates = options.GeminiOptions.SkipAndRemoveDups;
var dataSourceFactory = new GeminiDataSourceFactory(options.StitcherOptions, options.GeminiOptions.GenomePath,
options.GeminiOptions.SkipAndRemoveDups, refId,
Path.Combine(outdir, "Regions.txt"), debug: options.GeminiOptions.Debug);
var dataOutputFactory = new GeminiDataOutputFactory(options.StitcherOptions.NumThreads);
var samtoolsWrapper = new SamtoolsWrapper(options.GeminiOptions.SamtoolsPath, options.GeminiOptions.IsWeirdSamtools);
var geminiWorkflow = new GeminiWorkflow(dataSourceFactory, dataOutputFactory,
options.GeminiOptions, geminiSampleOptions, options.RealignmentOptions, options.StitcherOptions, options.OutputDirectory, options.RealignmentAssessmentOptions, options.IndelFilteringOptions, samtoolsWrapper);
Directory.CreateDirectory(outdir);
geminiWorkflow.Execute();
//var logger = new Illumina.CG.Common.Logging.Logger(taskLogDir, $"GeminiTaskLog_{chrom}.txt");
//var task = _taskCreator.GetCliTask(cmdLineList.ToArray(), chrom, exePath, outdir, chromRefIds[chrom], logger,
// string.IsNullOrEmpty(_options.GeminiSampleOptions.IntermediateDir)
// ? null
// : Path.Combine(_options.GeminiSampleOptions.IntermediateDir, chrom));
//tasks.Add(task);
Console.WriteLine($"Completed Gemini Workflow for {chrom}");
var path = (Path.Combine(outdir, "merged.bam.sorted.bam"));
taskDirectories.Add(outdir);
//paths[refId] = path;
return(path);
}
/// <summary>
/// Runs a two-pass STAR alignment for a given set of RNA-Seq fastq files,
/// or it performs a Bowtie2 alignment for WGS or exome sequencing files.
/// </summary>
public void PerformAlignment()
{
int starThreads = Math.Min(18, Parameters.Threads); // 18 max, otherwise it throws a segmentation fault in sorting the BAM files
if (Parameters.ExperimentType == ExperimentType.RNASequencing)
{
// Alignment preparation
WrapperUtility.GenerateAndRunScript(WrapperUtility.GetAnalysisScriptPath(Parameters.AnalysisDirectory, "GenomeGenerate.bash"),
STARWrapper.GenerateGenomeIndex(
Parameters.SpritzDirectory,
Parameters.Threads,
Parameters.GenomeStarIndexDirectory,
new string[] { Parameters.ReorderedFastaPath },
Parameters.GeneModelGtfOrGffPath,
Parameters.Fastqs))
.WaitForExit();
// there's trouble with the number of open files for sorting and stuff, which increases with the number of threads
// 18 is the max that works with the default max number of open files
TwoPassAlignment(starThreads, Parameters.OverwriteStarAlignment);
}
else
{
foreach (string[] fastq in Parameters.Fastqs)
{
SkewerWrapper.Trim(Parameters.SpritzDirectory, Parameters.AnalysisDirectory, Parameters.Threads, 19, fastq, false, out string[] trimmedFastqs, out string skewerLog);
FastqsForAlignment.Add(trimmedFastqs);
}
TopHatWrapper.GenerateBowtieIndex(Parameters.SpritzDirectory, Parameters.AnalysisDirectory, Parameters.ReorderedFastaPath, out string bowtieIndexPrefix);
List <string> alignmentCommands = new List <string> {
"echo \"Aligning reads with bowtie2.\""
};
foreach (string[] fastq in FastqsForAlignment)
{
// alignment
alignmentCommands.AddRange(TopHatWrapper.Bowtie2Align(Parameters.SpritzDirectory, Parameters.AnalysisDirectory,
bowtieIndexPrefix, Parameters.Threads, fastq, Parameters.StrandSpecific, out string sortedBamPath));
alignmentCommands.Add(SamtoolsWrapper.IndexBamCommand(sortedBamPath));
// mark duplicates
GATKWrapper gatk = new GATKWrapper(1);
alignmentCommands.AddRange(gatk.PrepareBamAndFasta(Parameters.SpritzDirectory, Parameters.AnalysisDirectory, Parameters.Threads, sortedBamPath, Parameters.ReorderedFastaPath, Parameters.Reference));
alignmentCommands.Add(SamtoolsWrapper.IndexBamCommand(gatk.PreparedBamPath));
SortedBamFiles.Add(sortedBamPath);
DedupedBamFiles.Add(gatk.PreparedBamPath);
}
WrapperUtility.GenerateAndRunScript(WrapperUtility.GetAnalysisScriptPath(Parameters.AnalysisDirectory, "BowtieAlignment.bash"), alignmentCommands).WaitForExit();
}
}
//wrapper should now handle all throwing and catching..
protected override void ProgramExecution()
{
_options.GeminiSampleOptions.InputBam = _options.InputBam;
_options.GeminiSampleOptions.OutputFolder = _options.OutputDirectory;
_options.GeminiSampleOptions.OutputBam = Path.Combine(_options.OutputDirectory, "out.bam");
_options.GeminiOptions.Debug = _options.StitcherOptions.Debug;
// Gemini defaults different than stitcher defaults
_options.StitcherOptions.NifyUnstitchablePairs = false;
// Set stitcher pair-filter-level duplicate filtering if skip and remove dups, to save time
_options.StitcherOptions.FilterDuplicates = _options.GeminiOptions.SkipAndRemoveDups;
var dataSourceFactory = new GeminiDataSourceFactory(_options.StitcherOptions, _options.GeminiOptions.GenomePath,
_options.GeminiOptions.SkipAndRemoveDups, _options.GeminiSampleOptions.RefId, Path.Combine(_options.OutputDirectory, "Regions.txt"), debug: _options.GeminiOptions.Debug);
var dataOutputFactory = new GeminiDataOutputFactory(_options.StitcherOptions.NumThreads);
var samtoolsWrapper = new SamtoolsWrapper(_options.GeminiOptions.SamtoolsPath, _options.GeminiOptions.IsWeirdSamtools);
var geminiWorkflow = new GeminiWorkflow(dataSourceFactory, dataOutputFactory, _options.GeminiOptions,
_options.GeminiSampleOptions, _options.RealignmentOptions, _options.StitcherOptions, _options.OutputDirectory, _options.RealignmentAssessmentOptions, _options.IndelFilteringOptions, samtoolsWrapper);
geminiWorkflow.Execute();
}