mapping initial libraries to Trinity assembly

Question

[lay_0]

I am worrying about something I came up while writing down my methods:

I assembled a transcriptome for a non-model species using the trinity pipeline. Adapters were removed at the sequencing center so I didn't have to worry about that. I checked at the quality of the libraries with fastqc and decided to trim some initial low quality bases and some trailing bases, retaining reads >85 bases. I did the trimming in the two concatenated files (one containing all the right reads and other all the left reads for each paired end library ), and used that to assemble my transcriptome.

Then I used RSEM to quantify my libraries against the assembly, and for this I used the initial 24 files (12 per paired end side). I now just realized that these initial files were not trimmed, trimming was done in the concatenated files used for de novo assembly..., could that affect somehow the library quantification? and as a consequence the differential expression?.

I can't figure out how many sequences were trimmed-out, I can see, however how many were retained (98.7%), after trimming...

can someone give me some input?. I worry I have to repeat the RSEM and all the following steps...

thank you!

[st.ph.n]

did you run trimmomatic within the trinity assembly pipeline? Does it not output trimmed files?

[lay_0]

I run trimmomatic separately, before starting Trinity, I may have missed some output there :(

[st.ph.n]

And you use the trimmomatic output files for the assembly, not the raw files? If so, you should be using the files yo u used to make the assembly for RSEM. Remember if you have multiple treatments, and are going to run downstreem DE analysis, align the trimmed reads per sample separately to your de novo transcriptome. Then proceed to downstream DE analysis with the DE program of your choice.

[lay_0]

yes, trimmomatic output was used for the assembly. thanks, that is what I suspected....i need to repeat the RSEM. I was hoping some one would say it didn't matter. If my reads were already trimmed for adapters, and they are of really good quality before trimming... do you think is still a problem using those to run RSEM in the trinity assembly?

[GenoMax]

Use ADD REPLY/ADD COMMENT to provide additional information. Don't SUBMIT ANSWER unless it is a new answer for your original question.

[st.ph.n]

If you ran RSEM separately on treatments/samples, I suspect you should be ok.