CostaLab / reg-gen

Regulatory Genomics Toolbox: Python library and set of tools for the integrative analysis of high throughput regulatory genomics data.
https://reg-gen.readthedocs.io/
Other
106 stars 30 forks source link

TRIPLEXES vs Triplexator #186

Open chiaracicco opened 3 years ago

chiaracicco commented 3 years ago

Hi, I was trying to apply TRIPLEXES to my data running rgt-TDF get_TTS but I could not find some common Triplexator's parameters, e.g. -L and -E.

Instead, these parameters are available to use in rgt-TDF region-test.

I briefly read the code and I was wondering if region-test usually runs Triplexator and if TRIPLEXES in general executes Triplexator.

Thank you for your help. Best, Chiara

chaochungkuo commented 3 years ago

Hi Chiara,

TRIPLEXES is very similar to Triplexator and their result should be exactly the same. Below is the help message of get_TTS:

Subparser 'get_TTS'
usage: rgt-TDF get_TTS [-h] [-i  ] [-tts  ] [-r  ] [-organism  ] [-l  ] [-e  ]
                       [-c  ] [-fr  ] [-fm  ] [-of  ] [-mf] [-rm  ]

optional arguments:
  -h, --help    show this help message and exit
  -i            Input BED file of the target regions
  -tts          Output BED file of the TTSs
  -r            Input FASTA file of the RNA
  -organism     Define the organism
  -l            [Triplexes] Define the minimum length of triplex (default: 20)
  -e            [Triplexes] Set the maximal error-rate in % tolerated
                (default: 20)
  -c            [Triplexes] Sets the tolerated number of consecutive errors
                with respect to the canonical triplex rules as such were found
                to greatly destabilize triplexes in vitro (default: 2)
  -fr           [Triplexes] Activates the filtering of low complexity regions
                and repeats in the sequence data (default: off)
  -fm           [Triplexes] Method to quickly discard non-hits (default:
                0).'0' = greedy approach; '1' = q-gram filtering.
  -of           [Triplexes] Define output formats of Triplexes (default: 1)
  -mf           [Triplexes] Merge overlapping features into a cluster and
                report the spanning region.
  -rm           [Triplexes] Set the multiprocessing

As you can see, there are many parameters available for TRIPLEXES, including -l and -e.

Please let me know if you need further help.

Joseph

chiaracicco commented 3 years ago

Dear @chaochungkuo, we noticed the -l and -e, what we missed are -E and -L.

In particular -L is crucial for some of the lncRNA we are studying. Without -L we are forced to use triplexator instead of triplexes.

There is something not clear for us, indeed seems that -L can be passed to triplexes when it is called trough region-test, e.g. rgt-TDF regiontest -r RNA.fa -bed ChIRP.bed -rn RNA -organism mm10 -o . -n 1000 -l 10 -e 20 -par L_-1_-E_-1.

Thank you Chiara

igcf commented 3 years ago

Dear Chiara,

Acording to http://bioinformatics.org.au/tools/triplexator/manual.html

-e is the maximun error rate -E is the maximum number of errors

Triplexes only provide the parameter -e and if you execute Triplexator without -E the outputs are the same.

Can you indicate the meaning of -L in triplexator? We cannot find this parameter in their manual.

Note also that Triplexator and Triplexes predict the same binding sites if the same parameterization, but their computational time and memory use differ. Also, triplexes provide distinct outputs (bed file like) which are easier to handle when comparing to common outputs of sequencing/epigenetic experiments.

I hope this helps.

best, Ivan

Em ter., 25 de mai. de 2021 às 11:43, chiaracicco @.***> escreveu:

Hi, I was trying to apply TRIPLEXES to my data running rgt-TDF get_TTS but I could not find some common Triplexator's parameters, e.g. -L and -E.

Instead, these parameters are available to use in rgt-TDF region-test.

I briefly read the code and I was wondering if region-test usually runs Triplexator and if TRIPLEXES in general executes Triplexator.

Thank you for your help. Best, Chiara

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/CostaLab/reg-gen/issues/186, or unsubscribe https://github.com/notifications/unsubscribe-auth/AC4XAAW2LGMP6XTJWRE3D7DTPNWN7ANCNFSM45O6RGIA .

--

www.costalab.org Institute for Computational Genomics RWTH University Hospital Aachen