Adding Manta module, associated pipeline & config changes #27
Conversation
timothyjsanders
changed the title
| * **map-qual:** >= 20 (Applied / Parameterized) | ||
| * **pe:** >= 5 (Not yet Applied / Non-parameterized) | ||
| * **sr:** >= 5 (Not yet Applied / Non-parameterized) | ||
| * **keep_imprecise:** >= true (Not yet Applied / Non-parameterized) | ||
|
|
||
|
|
There was a problem hiding this comment.
Maybe we want to say default instead of Non-parameterized?
There was a problem hiding this comment.
Maybe something like the following? Since there are default params in Delly, but some the user can't change.
**map-qual:** >= 20 (Default / Parameterized)
**pe:** >= 5 (Default / Non-parameterized)
**sr:** >= 5 (Default / Non-parameterized)
There was a problem hiding this comment.
Hmm then we should add the Delly default values? I meant the Delly default values not our default. Also, PE and SR are not independent values here. You can execute delly call, etc to see avilable options.
There was a problem hiding this comment.
Sorry I think I wasn't clear earlier. So a parameter that is listed as "Default / Non-parameterized" (for example pe >= 5) is a Delly default that isn't parameterized in the call-gSV pipeline. The map-qual parameter, on the other hand, has a default value of 20 and is parameterized in the call-gSV pipeline.
Our original intention was to do as you suggest, include the Delly default values, but also to indicate which ones users were able to change. It does seem like our list might be out of date though
There was a problem hiding this comment.
Yup, so maybe we can create a table to show - Delly default, call-gSV default? This is related to #6 . This PR is to add Manta so this can be done in a separate PR. Btw, do you know if Manta has the same option to set a PE/SR cut-off?
There was a problem hiding this comment.
Manta doesn't have the same options that Delly does (or at least I don't know where to find them). Perhaps this would be a filtering step after Manta runs? I've added a table in the README with the Delly defaults and whether or not they are available to set in the pipeline.
README.md
Outdated
|
|
||
| Currently the following filters are applied and or considered for application and parameterization in subsequent releases: | ||
| <br> | ||
| **Delly** | ||
| * **map-qual:** >= 20 (Applied / Parameterized) | ||
| * **pe:** >= 5 (Not yet Applied / Non-parameterized) | ||
| * **sr:** >= 5 (Not yet Applied / Non-parameterized) | ||
| * **keep_imprecise:** >= true (Not yet Applied / Non-parameterized) |
There was a problem hiding this comment.
I think Delly keeps imprecise sites as default. So, it's already applied as default.
There was a problem hiding this comment.
Got it, I can clarify some of the wording in the readme to make it more clear what's applied and what the user can modify
README.md
Outdated
| * **pe:** >= 5 (Not yet Applied / Non-parameterized) | ||
| * **sr:** >= 5 (Not yet Applied / Non-parameterized) |
There was a problem hiding this comment.
Required PE/SR clique size is 2 as default (PE/SR evidence combined) -z [ --min-clique-size ] arg (=2) min. PE/SR clique size - we may want to increase this default value as we used to do before.
Initial call-gSV runs will be used to detect overall gSV candidates so we could go with stringent cut-offs since our germline cohort is pretty large and then the 2nd discovery (regenotype-gSNP) can be more permissive. Any thoughts? @jmlivingstone @pboutros
There was a problem hiding this comment.
I think we'll have use-cases where we want to just run Manta and skip the delly and regenotyping steps (for compute efficiency), so I'd suggest not making the initial discovery too permissive here.
There was a problem hiding this comment.
Yeah, that's true. For regenotyping using Delly, we could add confident SV candidates predicted by Manta but I'm not sure if it's easy to incorporate.
There was a problem hiding this comment.
Should we parameterize all of the delly call and delly cnv parameters in another PR? We have just a handful of them available for modification by users currently.
There was a problem hiding this comment.
I don't think we need to parameterize everything - Caden resolved this issue by allowing extra arguments. See https://github.com/uclahs-cds/pipeline-call-sSNV/blob/master/pipeline/modules/mutect2-processes.nf
README.md
Outdated
| | Parameter | Delly default | call-gSV default | Description | | ||
| |:------------|:----------|:-------------------------|-------------| | ||
| | `svtype` | ALL | | SV type to compute (DEL, INS, DUP, INV, BND, ALL) | | ||
| | `map-qual` | 20 | 20 | Minimum paired-end (PE) mapping quality | |
There was a problem hiding this comment.
Looks like Delly default is not 20
-q [ --map-qual ] arg (=1) min. paired-end (PE) mapping quality
There was a problem hiding this comment.
Good catch, thanks!
| @@ -61,13 +61,22 @@ A directed acyclic graph of your pipeline. | |||
|
|
|||
There was a problem hiding this comment.
I just noticed that the flow diagram looks outdated?
There was a problem hiding this comment.
I will update the pipeline in another PR shortly
See outputs in /hot/pipelines/development/slurm/call-gSV/outputs/v3-manta/amini/TWGSAMIN000001-T001-S01