all RdRP+ macro/micro contigs

[rchikhi]

This issues describes the procedure to search all of our contigs against RdRP and presents results.
(Slack thread: https://hackseq-rna.slack.com/archives/C012H9SDQCA/p1615948152031200)

Input: FASTA files of contigs, either assembled using micro (all SRA .pro DIAMOND hits assembled with rnaviralspades) or macro (all from s3://lovelywater/assembly/contigs/, i.e. all CoV + dicistro + quenya + satellite + 1k random subset assembled using either coronaSPAdes or rnaviralspades).

Output: FASTA of all the contigs that hit RdRP either with HMM and/or palmscan, i.e. the RdRP+ contigs:
s3://serratus-rayan/pro-assembly/rdrpplus.micro.fa
s3://serratus-rayan/pro-assembly/rdrpplus.macro.fa
total size: 8.2 GB

hmmsearch was run using an exhaustive collection of RdRP HMMs:
https://gitlab.pasteur.fr/rchikhi_pasteur/serratus-rdrp-analysis/-/blob/master/hmm_macro_micro/RdRP_all.v2.hmm

alignments were made using this script:
https://gitlab.pasteur.fr/rchikhi_pasteur/serratus-rdrp-analysis/-/blob/master/hmm_macro_micro/align_hmm_to_contigs.sh

[rchikhi]

Some stats:

• number of macro RdRP+ contigs: 6,822,262
  total size: 5,828,043,099 bp
  longest macro RdRP+ contig: 1,086,412 bp
  N50: 1,870 bp

• number of micro RdRP+ contigs: 4,631,850
  total size : 2,158,973,751 bp
  longest micro RdRP+ contig: 16,630 bp
  N50 : 622 bp