NAME

fasda abundance - Compute abundances from SAM/BAM/CRAM data

SYNOPSIS

fasda abundance \


    --debug \


    --show-gene-name \


    --ignore-chromosome-order \


    --feature-type extended-RE \


    --output-dir dir \


    features.gff3 alignment-file [alignment-file ...]

DESCRIPTION

fasda abundance computes raw read counts from one or more SAM/BAM/CRAM files overlapping each features in a GFF3 file. Output is an abundance.tsv file in the same format as those produced by kallisto. Estimated counts may not match those produced by kallisto as they are calculated differently. However, the ratio of kallisto counts / fasda counts should be fairly consistent, and the fold-changes produced by fasda fold-change should therefore be highly similar.

The GFF3 and SAM/BAM/CRAM inputs must be sorted in the same order. By default, sort order is checked for both input streams using bl_chrom_name_cmp(3), and fasda abundance aborts if either is out of order. If this check is disabled using --ignore-chromosome-order, and chromosomes in the inputs are not in the same order, fasda abundance may produce incorrect output.

Abundances in the output are raw counts of properly paired reads that aligned to a feature of the selected type, are not flagged as secondary, duplicate, or supplementary alignments, and did not fail QC. For paired-end reads, both mates are counted separately, so the counts should be expected to reflect about double the number of fragments aligned.

Any normalization of counts is left to downstream tools such as fasda-normalize(1).

OPTIONS

--debug

Turn on debugging output

--show-gene-name

Show gene name rather than feature (mRNA/transcript) name in output.

--ignore-chromosome-order

Do not abort if GFF3 or SAM/BAM/CRAM input chromosomes are not sorted in order determined by bl_chrom_name_cmp(3). If this option is used, you must ensure that the order of the chromosomes is the same in the GFF3 and SAM/BAM/CRAM input.

--feature-type extended-RE

Indicate which feature types (column 3) in the GFF3 input each read should be checked for overlap with. This is an extended regular expression (see regex(3)). The default is "RNA$|transcript$|gene_segment$", i.e. any feature ending in "RNA", "transcript", or "gene_segment", such as "mRNA", "snRNA", "transcript", "unconfirmed_transcript", etc. For Ensembl GFFs, this matches features of type "transcript" in the GTF, which is the default for some other tools, including kallisto and gffread.

--output-dir directory

Indicate where -abundance.tsv output files should go. The default is the same directory is the SAM/BAM/CRAM input.

FILES

features.gff3
alignment.bam

SEE ALSO

fasda-normalize(1), fasda-fold-change(1)

BUGS

Please report bugs to the author and send patches in unified diff format. (man diff for more information)

AUTHOR

J. Bacon