what is the normal range of htseq-count output for homo sapiens RNA-Seq

Question

[Morteza]

I used hisat2 for mapping reads to reference genome and htseq-count for counting features. I downloaded reference genome sequence and annotation files from ensemble ftp portal. The output of ht-seq count is as follows:

__no_feature 3,324,819

__ambiguous 2,000,840

__too_low_aQual 937,875

__not_aligned 243,839

__alignment_not_unique 5,080,810

Is it in normal range?

[Michael]

I don't think it is possible to say on a general basis because experiments and protocols can be very different. Is this bulk or single-cell? You haven't given the total number of reads or how many reads overlap properly with a gene. The raw numbers are not informative, rather give percentages. My recommendation is to do a proper QC first, then decide eventual further steps based on the output.

[Morteza]

It is for bulk RNA-Seq and this is HISAT2 summary stats:

Total pairs: 42073612
    Aligned concordantly or discordantly 0 time: 1316719 (3.13%)
    Aligned concordantly 1 time: 38900504 (92.46%)
    Aligned concordantly >1 times: 1623109 (3.86%)
    Aligned discordantly 1 time: 233280 (0.55%)
Total unpaired reads: 2633438
    Aligned 0 time: 1462043 (55.52%)
    Aligned 1 time: 1052703 (39.97%)
    Aligned >1 times: 118692 (4.51%)
Overall alignment rate: 98.26%

[Michael]

So, everything is fine I guess...

[BioinfGuru]

wow that's a good alignment rate. Wish mine did that lol!