Feature Counts output interpretation

Question

[AP]

Hello everyone,

I am trying to count the number of genes mapped to the genome with feature counts using following code

featureCounts -p -T 5 -t CDS -s 2 -g gene_id -a /home/aubayp/rnaseq/star/annotation_file/Fov89-1A.gtf -o SL264821_counts.txt /home/aubayp/rnaseq/star/star_output/SL264821_outAligned.out.bam

The output I got looks really different than I got through htseq-count

Geneid  Chr     Start   End     Strand  Length  /home/aubayp/rnaseq/star/star_output/SL264821_outAligned.out.bam
g1      tig00000038;tig00000038 1332;2349       2265;2596       -;-     1182    0
g2      tig00000038     3236    4255    -       1020    0
g3      tig00000038     4922    5962    -       1041    0
g4      tig00000038;tig00000038;tig00000038;tig00000038 6885;7155;7572;7972     7079;7436;7907;8871     +;+;+;+ 1713    0
g5      tig00000038;tig00000038;tig00000038     9340;9396;10066 9349;9486;11032 +;+;+   1068    0
g6      tig00000038;tig00000038;tig00000038     19755;20173;20792       20111;20723;22463       +;+;+   2580    0
g7      tig00000038     22572   24827   -       2256    0
g8      tig00000038     25070   25699   +       630     0
g9      tig00000038     26223   28211   +       1989    0
g10     tig00000154     29402   32980   -       3579    0
g11     tig00000154     33148   34539   -       1392    0
g12     tig00000154     47202   48203   -       1002    39
g13     tig00000154;tig00000154;tig00000154;tig00000154;tig00000154;tig00000154;tig00000154;tig00000154;tig00000154     50993;51147;51226;51365;51570;51657;51741;51923;52394   51091;51181;5131
1;51511;51607;51694;51875;52345;52421   +;+;+;+;+;+;+;+;+       1029    81
g14     tig00000154;tig00000154;tig00000154     57684;57927;58153       57840;58089;58768       -;-;-   936     12
g15     tig00000154     59571   60497   -       927     1

In htseq I had output with just two columns with gene ID and counts. Could you please help me interpret this result from feature counts.

Thank you,

Ambika

Answer 1

[h.mon]

The discussion on the thread Question: How to generate a count matrix with featurecounts will help you understand featureCounts output. Basically, it is a tab-separated file, and some of its columns are "comma-separated" fields, because featureCounts outputs one field per exon. Geneid is first column and counts is last (or n last columns, in case you read n bam files), you can use cut to get a file similar to HTSeq.

[AP]

H. mon

Thank you for the clarification. I will use awk instead, just because I am more familiar with awk. :)