[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: ePCR fetch

On 6/24/19 2:55 PM, Andreas Tille wrote:
Hi Saira,

Hi Andreas,
On Mon, Jun 24, 2019 at 02:20:29PM +0100, Saira Hussain wrote:
I pushed towards the master my python script to be used for fetching data
from the database in order to be further modified and used for testing of
the ePCR package.

I confirm your fetch_fasta.py script works (=creates a data file).  I've
added a #! line on top which is more convenient since you can call the
script directly.

That's great. I will edit the heading and add some more steps in that script.
I have one problem down the line, as ePCR isn't happy with my final
converted file but I am working on it.

Could you please specify your problem?  What command line are you
calling and what is the output of this command?  What error is
Finally! I solved my problem after two days of trying different things. The problem is that to do an re-PCR comparison there need to be generated a few different files (starting from fasta and converting it using different tools such as fahash/famap etc). One of the files need to be the input primer for the ePCR process (short single-stranded nucleic acid).

As you will see on the new files pushed my primer is the following:

Alas, the space between the sequences isn't a space (or spaces) as I expected but a TAB! This resulted in the error I was getting:

vagrant@contrib-testing:~/or-debian-19/epcr/debian/tests$ ../../re-PCR -S ginko.hash -n 5 -g 0 -d 100-300 -o rbcL5.rePCR rbcL-primer.txt
! Fatal: format error in file rbcL-primer.txt

Finally I went through the source and find the problem!

I am going to send more files for review till the end of the day and hopefully finalise this test after quite some effort and pain :) Hopefully the next one would be way faster :)

Meanwhile could you give me some feedback? I hope a Python script is fine as
it's quite comfortable for me to write Python.

This script is not yet the test for ePCR - it just *documents* how the
used data are obtained.  In other words: You can write your
documentation in whatever language you want (some comment in the header
what this script is doing and why the data are fetched that way would be
Could I write further tests in Python instead of sh or bash scripts are

In principle the tests can be written in any language but I admit I do
not see a big need for it.  In all tests written for Debian Med
maintained packages shell was perfectly sufficient and we have a very
frequently used template[1] where you usually simply add a program call.
I'd consider a Python script unnecessary overkill and you should provide
more reasons why you think in this case some higher level language would
be better.
Looking forward to your comments and suggestions,

Kind regards


[1] https://salsa.debian.org/med-team/community/package_template/blob/master/debian/tests/run-unit-test

Reply to: