Really correcting bases??

mourisl / Rcorrector

Error correction for Illumina RNA-seq reads

GNU General Public License v3.0

63 stars 18 forks source link

Really correcting bases?? #37

Open rafaeltiveron opened 1 year ago

rafaeltiveron commented 1 year ago

I would like to share my experience with Rcorrector. I've tested one sample in two conditions:

Corrected with Rcorrector (k 31)
Not corrected

Both them were aligned against RefSeq genome with STAR. But sample not corrected by Rcorrector had more aligned reads than corrected fq output file. So, someone can explain/suppose something to share with us??

mourisl commented 1 year ago

Sorry for the delayed reply. What is your read length, what was your running command, and how did you count the number of aligned reads? Thank you.

rafaeltiveron commented 1 year ago

There is using run_rcorrector.pl:

-k 31
read length 75, for single-end (-s)
36 threads (-t)
Kept default other remaining options

I've counted unmapped reads to 4 genomes (--outReadsUnmapped Fastx), output by STAR, which seems more when rcorrector is used, independently of the reference used.