Add nice MultiQC summary assigning hashes to genes

[olgabot]

I wanted to add it in this PR (#41) but that PR is already huge and it's pretty out of scope. It would require adding a lot more summarization/etc code and I don't even know what I really want to put here yet.

[olgabot]

Example using csvtk(#51) to filter the featureCounts output to only the genes with nonzero reads.

 Mon 18 May - 09:52  ~/data_lg/kmer-hashing/brawand2011/predictorthologs/molecule-protein_ksize-27_log2sketchsize-14/mini/featureCounts/gene_counts 
 olga@tesla  csvtk filter -t -f "8>0" $ID\_gene.featureCounts.txt
Geneid  Chr     Start   End     Strand  Length  gene_name       hash-3557262534756__SRR306777_GSM752631_mml_br_F_1__reads_in_shared_hashes.bam
ENSMMUG00000014809      7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7;7   82441976;82442067;82442067;82442121;82442131;82442250;82442250;82442250;82442652;82442652;82442652;82443791;82443791;82443791;82461637;82461655;82461655;82462295;82462295;82465578;82465578;82465698;82465703 82442107;82442150;82442150;82442150;82442150;82442353;82442410;82442410;82442765;82442765;82442765;82443948;82443948;82443948;82461702;82461702;82461702;82462409;82462409;82465818;82465818;82465818;82465818       -;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-;-   1030    HNRNPC  40

In my mini test dataset, the full combination of differential hashes plus all bams is ~3000 files:

(base)
 ✘  Mon 18 May - 10:10  ~/data_lg/kmer-hashing/brawand2011/predictorthologs/molecule-protein_ksize-27_log2sketchsize-14/mini/featureCounts/gene_counts 
 olga@tesla  ls -1 | wc -l
3221

Each of these is ~12MB which is pretty big considering there's thousands of them.

(base)
 Mon 18 May - 10:11  ~/data_lg/kmer-hashing/brawand2011/predictorthologs/molecule-protein_ksize-27_log2sketchsize-14/mini/featureCounts/gene_counts 
 olga@tesla  ll |head
Permissions Size User Group Date Modified Name
.rw-r--r--@  12M olga czb   15 May  2:26  hash-3557262534756__SRR306777_GSM752631_mml_br_F_1_gene.featureCounts.txt
.rw-r--r--@  12M olga czb   15 May  2:27  hash-3557262534756__SRR306778_GSM752632_mml_br_M_1_gene.featureCounts.txt
.rw-r--r--@  12M olga czb   15 May  2:31  hash-3557262534756__SRR306786_GSM752640_mml_lv_F_1_gene.featureCounts.txt
.rw-r--r--@  12M olga czb   15 May  2:28  hash-3557262534756__SRR306787_GSM752641_mml_lv_M_1_gene.featureCounts.txt
.rw-r--r--@  11M olga czb   15 May  2:37  hash-3557262534756__SRR306827_GSM752680_ppa_br_F_2_gene.featureCounts.txt
.rw-r--r--@  11M olga czb   15 May  2:31  hash-3557262534756__SRR306836_GSM752689_ppa_lv_M_1_gene.featureCounts.txt
.rw-r--r--@  12M olga czb   15 May  2:46  hash-185985883680630__SRR306777_GSM752631_mml_br_F_1_gene.featureCounts.txt
.rw-r--r--@  12M olga czb   15 May  2:44  hash-185985883680630__SRR306778_GSM752632_mml_br_M_1_gene.featureCounts.txt
.rw-r--r--@  11M olga czb   15 May  2:49  hash-185985883680630__SRR306827_GSM752680_ppa_br_F_2_gene.featureCounts.txt

I don't see a use case where someone would want ALL of the featureCounts output, especially since so many of the genes are undetected. Maybe I'm wrong?? But I think the best way for summarizing is to do csvtk filter within the featureCounts process and remove all zero-count genes.