Incorrect input data for srna tutorial

[hexylena]

Hi @malloryfreeberg

One of the .eu users ran the Differential abundance testing of small RNAs and encountered some issues with the tutorial. :/

In the Hands on after FastQC it is said that the datasets should have adapters (picture attached from the hands-on). But the output we got shows no adapters and also TrimGalore was unable to trim reads?

@wm75 reproduced the report and noted that the data uploaded to zenodo is already trimmed. Do you potentially have the dataset for the step before that? Otherwise we'll have to find someone to rewrite the tutorial to remove this step.

Lastly a minor issue, the "Zenodo" link in here points to the wrong record.

[hexylena]

Hey @bebatut or @mblue9, I've received no response to this (her or on gitter), so since you're the other two maintainers for transcriptomics, would either of you have suggestions for how to handle this? In the absence of someone to fix it, should we open a PR to disable this tutorial until it can be fixed?

[vebaev]

Dear all,
please keep the tutorial as it is for the next week as we will have Galaxy Training - we will just skip the step of TrimGalore as the data is trimmed! Please do not make any changes as we already prepare to make this tutorial as it is in front of the workshop trainees.
After the event, I can help to update it as we found also another things to fix! But please wait for now! I will write after the event!

[hexylena]

@vebaev we have archives of old versions of the training material here as well, in case any changes are ever made before a training. https://training.galaxyproject.org/archive/

If you're able to contribute a fix, that's great! Thank you!

[vebaev]

@erasche @bebatut @mblue9
Thanks after the event (around 3-4 April) I will try to fix it and also correct some other stuff in it!

[malloryfreeberg]

Hi @erasche! Sorry, just getting to this now. Let me take a look at tutorial and materials. I'm fairly sure I used published data, so it shouldn't be difficult to get the raw data files again.

[hexylena]

Hi @malloryfreeberg, no problem! Ok, brilliant, if we can get the original data that will be ideal. Thanks for answering and looking into it! :)

[hexylena]

@vebaev do you have any more updates for this? I think I got an email from one of your colleagues and during helping them I identified more issues that we should fix in this training.

• we should use name tags on DCs to make this more manageable
• piRNAs are extracted and never used/discussed further
• DESeq2 step uses incorrect inputs I think? (factor symplekin needs input symplekin, factor "RNAi" should maybe be renamed to "blank" and needs blank inputs, not symplekin, or? Maybe I am not familiar enough with the study being done here, but it seems strange to compare the same dataset twice as two factors.)
• Maybe just remove the "Coming soon"
• Some of the inputs from zenodo are referred to without their name, using names like "the reference mRNA and TE fasta file" rather than dmel3_transcriptomics_.....fa.gz
• s/The read in unmapped/The read is unmapped/
• I had forgotten about this issue when I went to write more comments, but trimming seemed to work OK this time. @wm75 do you still have that history?

[vebaev]

We have a workshop tomorrow. Yes, my PostDoc contacted you :) We couldn't do it beyond Deseq2, since the version may be different (not sure). I have the history from before 2-3 months where only 1-2 groomers were missed, but now it seems more issues. I will see how will go tomorrow and will tell her to post here the outcome and communicate with you.

[hexylena]

@vebaev I did get it working eventually! Here is the history I sent to your Post Doc: https://usegalaxy.eu/u/helena-rasche/h/transcriptomicssrna

[vebaev]

@erasche many thanks, this will help for tomorrow!