how to cut the adapter without knowing the seq?

[Noah Pieta]

i'm a green hand in RNA-seq data analysis. i have a raw data that need to cut the adapters. i'm using the cutadapter software to do the job. but i don't know the adapter seq. what can i do to solve the problem.
thank you very much!

[Wallysb01]

You could run fastqc and see if there are any overrepressented adapter sequences. It will identify which adapter or primer sequences are in there, if it finds too many of them.

[simonandrews]

If you're using Illumina sequences then you can trim with the common Illumina prefix which is present at the start of pretty much all of their adapters. If you use our trim_galore trimmer (which is a wrapper around cutadapt) then it will do this for you automatically.

[mgogol]

I find the kmer plots of fastqc useful for this. If there are adapters at the end or something, you'll see the corresponding kmers coming up at the end. You can usually use this plus the contaminants list from fastqc to figure out what to trim.

I'm more familiar with trimming for small RNA data though. I don't usually trim anything from RNA-seq.

[vineet jha]

Sangenix

Sangenix can perform adapter trimming , from fastqc report. so no need to provide adapter seq for trimming. and it is all automated.

Try Sangenix: http://www.sangenix.com/

[simonandrews]

Sangenix can perform adapter trimming , from fastqc report. so no need to provide adapter seq for trimming. and it is all automated.

Try Sangenix: http://www.sangenix.com/

How does it decide what to trim based on the FastQC report? I'm not sure I could reliably figure that out manually from just the information FastQC provides.

[vineet jha]

Sangenix Trim Adapter from fastQC report

Sangenix Reads the fastQC report , and create a list of over represented sequences . It uses these Sequence as adapter trimming from corresponding fastQ file . it trims the bases based upon user set parameter also. the attached figure show , before and after images of adapter trimming after QC and filtering step done by sangenix

http://www.sangenix.com/

[Noah Pieta]

If you're using Illumina sequences then you can trim with the common Illumina prefix which is present at the start of pretty much all of their adapters. If you use our trim_galore trimmer (which is a wrapper around cutadapt) then it will do this for you automatically.

thx very much！ i have tried it

[sklages]

Sangenix can perform adapter trimming , from fastqc report. so no need to provide adapter seq for trimming. and it is all automated.

Try Sangenix: http://www.sangenix.com/

Well, .. and were can the software be downloaded? I could not find any download link!?

[simonandrews]

Sangenix Reads the fastQC report , and create a list of over represented sequences . It uses these Sequence as adapter trimming from corresponding fastQ file . it trims the bases based upon user set parameter also. the attached figure show , before and after images of adapter trimming after QC and filtering step done by sangenix

I'm not sure that's a great idea. The overrepresented sequences identification will really only spot if you have adapter dimers or other concatamers. If you have more normal read-through adapter contamination then these won't show up as the starting sequences will be more diverse. Also if you do get dimers they don't always show the full length of the adapter sequence to trim. If nothing else you'll normally find they're missing the initial 'A' at the start of the adapter (from the overhang ligation) so you'd need to add that, but we've also seen other truncated forms turn up. You're also potentially open to hitting other sequences which are overrepresented in your library. Doing this on a smallRNA library for example would cause a real mess.

[kcchan]

Yea using overrepresented sequences as the adapter sounds like a horrible idea. If you tried it with an amplicon or mate pair library all of your data is going to get wiped away.

[Wallysb01]

Well, the idea would be to only use fastqc to find which adapters were used, then trim those. If fastqc is finding overrepresented sequences other than known adapters, you ignore it and don’t trim it.

[kcchan]

But even then it'll have to show up in the overrepresented sequences and is fully dependent on whether or not the contaminant list is up to date. If I recall correctly the default list doesn't have Nextera nor the new targeted RNA adapters.

[simonandrews]

But even then it'll have to show up in the overrepresented sequences and is fully dependent on whether or not the contaminant list is up to date. If I recall correctly the default list doesn't have Nextera nor the new targeted RNA adapters.

Further contributions gratefully received

You can also add in more sequences yourself to the contaminants.txt file.

[rhinoceros]

i'm a green hand in RNA-seq data analysis. i have a raw data that need to cut the adapters. i'm using the cutadapter software to do the job. but i don't know the adapter seq. what can i do to solve the problem.
thank you very much!

Why can't you ask from the person who gave you the data?