feature request - add a summary of seq counts at each point during run

[afmcc]

e.g. #demultuplex , #trim, #fasta , #non-redundant etc

[afmcc]

I have been putting some summary statistics together for the sequencing I have done and I noticed that there are some samples that have an unusually low number of sequences in their FastA file. After looking at the trimming files it looks like there was an issue translating the FastQ to FastA where the FastA file is incomplete. I have included a table of the samples for which I have picked up this inconsistency (not too many affected – just 10 out of 4000 samples). The “PropPassQC” column is usually ~94% and these are the samples with less than 90% and they all have the issue where the number of sequences in the FastA file is less than what the trim report says it should be (“Reads written (passing filters)”). Despite this error they still pass my QC threshold of having 100,000 reads.

I was wondering if we could add a test where we check whether the number of reads in the FastA is the same as the FastQ and generates an error if they are different?

SampleID LabID Library numDemultReads numReadsFastA PropPassQC
B74793 B74793 SQ0741 804577 636783 0.7914507
B74883 B74883 SQ0741 645206 172998 0.2681283
B74948 B74948 SQ0741 712765 179106 0.2512834
B70703 961228 SQ0881 567223 369528 0.6514686
B73471 961407 SQ0881 626250 513904 0.8206052
B72445 961442 SQ0881 721064 163343 0.2265305
B72278 961460 SQ0881 1120105 305083 0.27237
B72288 961832 SQ1019 840772 267774 0.3184859
B72371 961935 SQ1019 785259 356252 0.4536745
B73670 962246 SQ1020 661944 448017 0.6768201