You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In cases where an incorrect R code is present, this causes an error with eggd_batch. The process for handling this error is to manually remove the sample from the manifest and then rerun eggd_batch. Manually removing information from the manifest means there is potential for human error to remove information from multiple samples.
Adding an option to ignore samples in the manifest and not generate reports for them at that point, similar to -iexclude_samples used for CNV calling, would remove this risk of human error.
The text was updated successfully, but these errors were encountered:
Thanks for the suggesstion, I think when we first discussed requirements we decided against doing this, as we said the manifest should be fixed in Epic rather than partly running a batch file, although maybe I'm wrong / misremembering / this isn't the same. What example of incorrect codes is it you're referring to?
Thanks @Emma-MacK for suggesting this and thinking about introduction of errors that's really good.
I wonder as discussed by Jethro above whether instead of building it into batch we could add the process of removing a sample from a batch in epic and once the r code is fixed the remaining sample would just be run with a new epic batch?
@jethror1 this was related to recent CEN run, 240130_A01303_0329_BH2HWHDRX5. eggd_dias_batch initially failed due to one of the samples having an incorrect test code requested, R184.1. R184.2 is in the valid test codes and R184.3 is in the CNV codes.
I agree, being able to remove a sample from a batch in EPIC would be handy and also resolve this.
In cases where an incorrect R code is present, this causes an error with eggd_batch. The process for handling this error is to manually remove the sample from the manifest and then rerun eggd_batch. Manually removing information from the manifest means there is potential for human error to remove information from multiple samples.
Adding an option to ignore samples in the manifest and not generate reports for them at that point, similar to -iexclude_samples used for CNV calling, would remove this risk of human error.
The text was updated successfully, but these errors were encountered: