3

I want to use TRAL to annotate tandem repeats in the reference genome of Caenorhabditis elegans. For this, I need to install some external software, such as Phobos. I've downloaded Phobos and I am using it by typing the path to it's executable like this:

input/software/phobos-v3.3.12-mac/bin/phobos-mac-intel-64bit-start-from-command-line genome.fa output.txt

This works quite well.

In order for TRAL to find Phobos, I changed TRAL's configuration file as explained here.

I've replaced PHOBOS = phobos with PHOBOS = /Users/user/Documents/.../project/input/software/phobos-v3.3.12-mac/bin/phobos-mac-intel-64bit-start-from-command-line

Then, I followed the tutorial Run and parse de novo repeat detection software and typed the following commands in Python 3 in the Terminal:

import os
from tral.sequence import sequence
from tral.paths import PACKAGE_DIRECTORY

proteome_HIV = os.path.join(PACKAGE_DIRECTORY, "examples", "data", "HIV-1_388796.faa")
sequences_HIV = sequence.Sequence.create(file = proteome_HIV, input_format = 'fasta')

So far so good. But the following command creates an error:

tandem_repeats = sequences_HIV[0].detect(denovo = True, detection = {"detectors": ["Phobos"]})

Did not find Phobos result file in /var/folders/z2/rt6qxntd3tg5bjmqlqnm0mdm0000gn/T/tmpvc20y3i4/001/PHOBOS/phobos.o

Note that if I use PHOBOS instead of Phobos in the command, I get the following error:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/user/anaconda/envs/python3env/lib/python3.6/site-packages/tral/sequence/sequence.py", line 162, in detect
    **kwargs['detection'])[0]
  File "/Users/user/anaconda/envs/python3env/lib/python3.6/site-packages/tral/sequence/repeat_detection_run.py", line 962, in run_detector
    Detectors(detectors, sequence_type)
  File "/Users/user/anaconda/envs/python3env/lib/python3.6/site-packages/tral/sequence/repeat_detection_run.py", line 872, in Detectors
    FINDER_LIST.keys()))
Exception: Unknown TR detector supplied (Supplied: ['PHOBOS']. Known TR detectors: dict_keys(['HHrepID', 'Phobos', 'TRED', 'T-REKS', 'TRF', 'TRUST', 'XSTREAM']))

I am working on Mac OS X. I installed TRAL with pip install tral.

Biomagician
  • 2,459
  • 16
  • 30
  • 2
    I can reproduce this, so I think it might be a TRAL bug. See https://github.com/acg-team/tral/issues/12 – Quantum7 Jul 17 '18 at 11:25
  • Are you running on DNA or protein? In general, TRAL support for DNA is weaker than we'd like it. If you're searching exons, I might try performing the translation first and running in AA mode. – Quantum7 Jul 27 '18 at 13:35
  • Running on DNA. I am interested in DNA repeats. Mini-and microsatellites are unlikely to code for anything. If I run it on the translation, then the repeats are not bound to appear at the DNA level because of the genetic code degeneracy. – Biomagician Jul 28 '18 at 13:31
  • OK. Just make sure you review the results carefully when using TRAL on DNA, as the parameters were trained for AA. We have a student working on improving DNA repeats now, so this should get easier. – Quantum7 Jul 30 '18 at 18:28

1 Answers1

3

TRAL 0.3.5 has bugs running and parsing Phobos. You can follow progress towards a fix on this issue: https://github.com/acg-team/tral/issues/16

Until the issue is fixed, you should remove Phobos from the repeat detector list in your tral config file (~/.tral/config.ini) and use a different repeat detector.

[sequence]
    [[repeat_detection]]
        DNA = TRED, T-REKS, TRF, XSTREAM
Quantum7
  • 146
  • 3