HT-seq, No feature, Alignment not unique Output

Question

[TM]

Hi, I got the output from HT-seq as the following

__no_feature    3138617 2814862 3510969 3772998 
__ambiguous 965037  905683  1065946 768189  
__too_low_aQual  474675 425696  548364  330561  
__not_aligned   300633  246494  331364  211173  
__alignment_not_unique  1228830 1114723 1379571 1104518 

Is it acceptable to have no feature read 3-4millions while the overall alignment read is around 20-25 millions read ( 98% overall alignment rate using HISAT2)?

Answer 1

[yhoogstrate]

Is it acceptable to have no feature read 3-4millions while the overall alignment read is around 20-25 millions read ( 98% overall alignment rate using HISAT2)?

This does not sound very worrisome to me. FFPE is typically much worse. To answer your question you will need to figure out where in the genome these 3-4 million reads are located. Often these are ribosomal RNA's, located in human alignments on both one of the alternate loci and chr21 (if I am right). This is usually not a bad thing. It could also be that these 3-4 million reads are scattered out throughout the entire reference genome, which could be an indication for DNA contamination. I once experienced this when RNA was sequenced using a combined DNA/RNA kit, which I would consider a bad thing.

I would also recommend to inspect the alignments in combination with the used GTF/GFF3 file to track down where and why these 3 million reads were not included.