Add blast.ml to lib/bioinfo#38
Conversation
…an unique prefix name for results and recover them more easily
|
Ajout également des sélecteurs dans prokka.ml pour pouvoir réutiliser les fichiers de sortie |
pveber
left a comment
pveber
left a comment
There was a problem hiding this comment.
This is a great start! In addition to the comments below, here are two other important points:
- remove modifications from
prokkaas they are not related to the current PR - add an
mlifileBlast
|
|
||
| let db_name = "db" | ||
|
|
||
| let fastadb fa dbtype = |
There was a problem hiding this comment.
better call it makedb since it's the name of the tool
There was a problem hiding this comment.
also dbtype should be the first argument and named
| mkdir_p dest ; | ||
| cmd ~env "makeblastdb" [ | ||
| opt "-in" dep fa ; | ||
| opt "-dbtype" ident dbtype ; |
There was a problem hiding this comment.
the type of dbtype is Bistro.Template.t, which is really not convenient for a user. I'd much prefer have a polymorphic variant type here [nucl | prot]
There was a problem hiding this comment.
So we have to create the type ? In the .ml ? I don't really see how to call the function then.
Something like type polymorphic_variant = Prot | Nucl in .ml and call Blast.makedb ~dbtype:Prot ref_prot outside ? With a "reconversion" to string somewhere to call the command line.
(it's not exactly like that because it's not working)
There was a problem hiding this comment.
actually if you use polymorphic variants `Nucl and `Prot, you don't need to declare the type and it's also easier when calling the function. So yes, basically a call would be just like you say Blast.makedb ~dbtype:`Prot ref_prot, and yes you have to provide the conversion function that will take a variant and make an appropriate Bistro.Template.t out of it.
| cmd "blastn" ~env [ | ||
| opt "-db" ident (dep db // db_name) ; | ||
| opt "-query" dep query ; | ||
| opt "-out" ident (dest // out_name) ; |
There was a problem hiding this comment.
what's the motivation for creating a directory here? Couldn't we give dest to the -out option?
There was a problem hiding this comment.
Because it's already created in makedb ? Or do we have to create it in the main pipeline ?
There was a problem hiding this comment.
Well yes, the same could be said about the database, I think in both cases you don't really need to create a directory and put the resulting file in it. The resulting file should be put at the location dest directly I think.
There was a problem hiding this comment.
To be more precise in think your wrapper could be simplified like this:
opt "-db" dep db ;
opt "-out" ident dest ;of course the wrapper for makedb should be modified accordingly.
|
I tried to create two different PR for blast and prokka but it fails, I don't know how to do it. |
|
You should create one branch per topic, one for blast the other for the fix on prokka, push both branches to your github depo, and create a new pull request from each. |
…to have an unique prefix name for results and recover them more easily" This reverts commit 1f01944.
pveber
force-pushed
the
master
branch
3 times, most recently
from
fa83704 to
49c7a0e
Compare
2 participants
Bonjour,
Voici le script .ml pour blast. Il y a uniquement blastn et blastp pour le moment, et toutes les options n'ont pas été mises, je ne sais pas si il faut toutes les ajouter ?