Comments (11)
Alright, I tested this on patient samples.
here is the plot w/o freyja_update
which then uses a Freyja database that is part of the container (from March 2024 or so)
here is the same data and command but with freyja_update
and we can see that the update function does something (attention, the sample columns are not exactly the same)
Let's check the "interesting" samples (if I can see the colors correctly)
IMSSC2-174-2024-00065, barcode13 - assigned JN 1.48.1 (designated 15.04.2024)
- w/o freyja update: JN.1.16.1
- w/ freyja update: 50% JN 1.16.1 and 50% JN1.48.1
IMSSC2-99-2024-00026, barcode 12 - assigned JN 1.16.2 (designated 15.04.2024)
- w/o freyja update: JN.1.16
- w/ freyja update: 80% JN.1.16.2 and 20% JN.1.16
IMSSC2-63-2024-00057, barcode 18 - assigned KP 2.3 (designated 15.04.2024)
- w/o freyja update: KP.2
- w/ freyja update: KP.2.3
Summary
This works very well! And the update function also does it's job. Still, I think we should implement a more robust way where the used freyja reference db is also versionized and stored. But this can be done in another issue/PR.
from porecov.
Amazing results @hoelzer !
from porecov.
I'd vote for deprecating LCS, then. If we need LCS results in the near future for whatever reason, we'd still have the option to run it easily
from porecov.
In this context, the help mssg
--screen_reads Determines the Pangolineage of each individual read (takes time, needs --freyja and/or --lcs)
Is wrong. I am not 100% sure how LCS was doing this, but Freyja is not giving a lineage assignment for each read. Freyja is doing a deconvolution of the variant calls to estimate the proportion of lineages.
In this context, I am also wondering @DataSpott @replikation if we could not remove the screen_reads parameter?
Then, the user can just activate freyja or LCS with the parameter that is anyway available?
from porecov.
In this context, the help mssg
--screen_reads Determines the Pangolineage of each individual read (takes time, needs --freyja and/or --lcs)
Is wrong. I am not 100% sure how LCS was doing this, but Freyja is not giving a lineage assignment for each read. Freyja is doing a deconvolution of the variant calls to estimate the proportion of lineages.
In this context, I am also wondering @DataSpott @replikation if we could not remove the screen_reads parameter?
Then, the user can just activate freyja or LCS with the parameter that is anyway available?
agree on this screen_reads part to simplify. but i prefer descriptions rather than tool names. (e.g. --screen_reads Uses freya to.....)
maybe deprecate LCS?
from porecov.
In this context, the help mssg
--screen_reads Determines the Pangolineage of each individual read (takes time, needs --freyja and/or --lcs)
Is wrong. I am not 100% sure how LCS was doing this, but Freyja is not giving a lineage assignment for each read. Freyja is doing a deconvolution of the variant calls to estimate the proportion of lineages.
In this context, I am also wondering @DataSpott @replikation if we could not remove the screen_reads parameter?
Then, the user can just activate freyja or LCS with the parameter that is anyway available?agree on this screen_reads part to simplify. but i prefer descriptions rather than tool names. (e.g. --screen_reads Uses freya to.....) maybe deprecate LCA?
From my pov we could discontinue LCS. And then screen_reads would simply activate freyja.
@Krannich479 @MarieLataretu are you using LCS for anything? From my pov it's not well-updated and freyja would be anyway preferable
from porecov.
@Krannich479 @MarieLataretu are you using LCS for anything? From my pov it's not well-updated and freyja would be anyway preferable
We occasionally investigate the LCS results in special cases. But I agree that LCS is not up-to-date (even if we could update it)!
Maybe deprecation is a nice compromise? @replikation , it would be usable/available for some time, right?
from porecov.
yep
from porecov.
Agree!
from porecov.
Regarding your freja
results, do you mind re-running freja if I error-correct you the read pools with herro
? This insight would bolster my (dis-)trust in EC for variant calling.
from porecov.
@Krannich479 pleasure.
We could also then do a re-basecalling of the run with v5 Dorado SUP and see any differences.
from porecov.
Related Issues (20)
- add skip scorpio parameter to pangolin HOT 1
- Only calculate NanoPlot after read filtering step HOT 5
- Add new V5 ARTIC primer BED HOT 5
- Medaka step fails in the -profile fastq_test HOT 3
- summary_report.py fails HOT 7
- publish primersitereport from medaka output
- VarSkipV2b primer does not work as expected HOT 7
- Update Medaka to support R10.4.1 models HOT 14
- Update --help to list up-to-date primer schemes that are supported
- MinKNOW/Guppy update needs new model for R10.4.1 5 kHz HOT 6
- Warning when execution report and timeline already exists HOT 1
- The pipeline fails in artic_ncov_wf_artic_medaka HOT 7
- new pangolin table columns HOT 1
- CovarPlot fails w/ custom BED HOT 5
- Process retry in slurm profile HOT 1
- Publish VCF files HOT 2
- Update medaka to get the latest models HOT 1
- [Question] CI for Variant Calling HOT 5
- Singularity container execution of pangolin crashes with recent nextflow versions HOT 6
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from porecov.