| | 289 | == [RNA-seq only] Get global coverage profile across transcripts == |
| | 290 | |
| | 291 | Do reads come from across the length of a typical transcript, or is there 3' or 5' bias (where most reads come from one end of a typical transcript)? |
| | 292 | |
| | 293 | One way to look at this is with Picard's CollectRnaSeqMetrics tool |
| | 294 | {{{ |
| | 295 | # Usage: |
| | 296 | java -jar /usr/local/share/picard-tools/CollectRnaSeqMetrics.jar INPUT=bamFile REF_FLAT=refFlatFile STRAND_SPECIFICITY=NONE OUTPUT=outputFile REFERENCE_SEQUENCE=/path/to/genome.fa CHART_OUTPUT=output.pdf VALIDATION_STRINGENCY=SILENT |
| | 297 | # Example command |
| | 298 | java -jar /usr/local/share/picard-tools/CollectRnaSeqMetrics.jar INPUT=WT.bam REF_FLAT=/nfs/genomes/mouse_mm10_dec_11_no_random/anno/refFlat.txt STRAND_SPECIFICITY=NONE OUTPUT=QC_metrics/WT.RnaSeqMetrics.txt REFERENCE_SEQUENCE=/nfs/genomes/mouse_mm10_dec_11_no_random/fasta_whole_genome/mm10.fa CHART_OUTPUT=QC_metrics/WT.RnaSeqMetrics.pdf VALIDATION_STRINGENCY=SILENT |
| | 299 | |
| | 300 | }}} |
| | 301 | The VALIDATION_STRINGENCY=SILENT option will keep the program from crashing if it finds something unexpected. The default: VALIDATION_STRINGENCY=STRICT |
| | 302 | |
| | 303 | |
