bowtie2

[Ivano Legnini]

Hi Ricardo,

I managed to run pigx in the end, with decent results, although I now
will play a bit around to understand some issues with the data.

However, there are a few issues which I'd like to report:
1. the overall architecture of the pipeline is a bit unclear: e.g. I
find a Deseq2 output for STAR-aligned reads and one for Salmon, but also
a featurecounts output which I'm not entirely sure how was generated and
how is used further.
2. the following is very concerning: there is at least one entry in the
top 10 in the original STAR-Deseq2 tsv file which is not appearing in
the html repor, paragraph 3.1 (results table). ???
3. It would be great to be able to play with the alignment parameters
(at least I could not find an immediate way of doing it within the
pipeline): I have ~50% reads flagged by STAR as not aligned: too short,
which I am now trying to address by separately running STAR with
different parameters.
4. as I've reported before, the pipeline does not run if there's a
newline in the settings file, which took me a good day to spot as the
reason why it was not running -> it would be great to have more explicit
error messages.

That said, it is a great tool for people like myself who are not so good
at computational analyses and despite the issue I've encountered it was
extremely useful for me.

All the best and thanks a lot for all the help

Ivano

--
Ivano Legnini, PhD

Postdoctoral fellow
Systems Biology of Gene Regulatory Elements
Berlin Institute for Medical Systems Biology
Max Delbrück Center for Molecular Medicine
in the Helmholtz Association
Hannoversche str. 28
10115 Berlin, Germany