Hi All, I just submitted a manuscript for metatranscripomic analysis results based on one of the SRA study. I was reporting some of the sequences from Trinity assembly for this study. I just got some comments from the reviewers saying that I missed to report some additional contigs. One of the reviewers was able to find additional contigs that I was not able to find in Trinity assembled database. I then alligned the fastq files (paired end) to the contigs reviewers found. I found that it has less than three hundred reads aligned. I could not find those contigs in my Trinity assembled database, but I was still able to align fastq reads (albeit less than 300 reads) to those contigs. Therefore, I was wondering if Trinity has failed to report those contigs with low coverage. Does Trinity drop the contigs with low coverge during de-novo assembly? Would anyone please help me understand why it was not reported by Trinity so I could write my rebuttal to them?
I think it's indeed likely that Trinity drops out low coverage contig during assembly because it will consider it as 'noise' .
I wonder though, where the reviewer got those contigs from? Did he assemble the data himself?
Yes, the reviewer assembled the data and probably using different assembly tool.
Some suggestions:
Ask him about his methods, otherwise you can't say much about the differences found.
Filter the reads that map to those contigs, and map them to your assembly.
When you map the reads used to assemble the transcriptome, what is the mapping rate? is it 95% or above?