Running pipeline for fungi data analysis

[floriantilliet]

Thank you for making the Global Soil Mycobiome pipeline publicly available.
I'm currently trying to run your pipeline on data from the JRC for analysing of fungal communities. I have a few questions about scripts 02.Extract_ITS.sh and 04.Prepare_UNITE_data.sh

In script 02.Extract_ITS.sh: first I am not sure about how removing sequences with incorrect primer orientation works. Don't the -g flag in cutadapt only search primers in the 3' direction ? Also, I am not sure about the purpose of each cutadapt command, don't the final ITSx command try to achieve the same result as the previous cutadapt commands ?

In Script 04.Prepare_UNITE_data.sh: the script mentions "untrimmed potentially partial sequences." I'm not sure about what they are exactly nor where they come from. Could you please indicate at which step this file should be generated?

I’ve carefullty looked at the paper and the repository documentation but would greatly appreciate your insights to ensure I’m using the pipeline correctly. Thank you for your time !

Best regards,
Florian Tilliet

[vmikk]

Hello Florian,

Thank you for your interest in the GSMс pipeline and for your thoughtful questions.

First, for the analysis of new data, I would recommend using our updated workflow - NextITS.
This updated pipeline builds on our previous GSMc (v1) approach with multiple significant improvements. For example, we redesigned primer checking step, added homopolymer error correction, etc.

not sure about how removing sequences with incorrect primer orientation works

The purpose of using several cutadapt commands was to remove artificial sequences that likely arise from errors in generating circular consensus sequences. For example, some sequences may contain multiple primer occurrences, where cutadapt trims only the "outer" primer, leaving an erroneous sequence. In other cases, primers in the wrong orientation (e.g., forward primer followed by reverse primer instead of the reverse complement of the reverse primer) indicate incorrect amplicons, which were also filtered out.

I am not sure about the purpose of each cutadapt command, don't the final ITSx command try to achieve the same result as the previous cutadapt commands

In our experience, removing the primers prior to ITS extraction led to a better overall performance (and improved the efficiency of ITS extraction, even though it might seem that ITSx alone could handle the task).

With kind regards,
Vladimir