Failure placementReduced.jplace has no placements

[sarah872]

Hi, I was running eukcc on an illumina assembly of a big eukaryotic genome and got the following error:

0/05/2020 15:25:11:  Starting EukCC
10/05/2020 15:34:36:  Running GeneMark-ES
10/05/2020 21:04:04:  Extracting protein locations
10/05/2020 21:04:05:  Searching for proteins to place in the tree
10/05/2020 21:04:09:  Processing Hmmer results
10/05/2020 21:04:16:  Placing proteins in tree
Uncaught exception: Failure("eukcc_wgs-holobiont/workfiles/pplacer/placementReduced.jplace has no placements!")
Fatal error: exception Failure("eukcc_wgs-holobiont/workfiles/pplacer/placementReduced.jplace has no placements!")
10/05/2020 21:06:49:  an error occured while executing guppy
10/05/2020 21:06:49:  No estimates were written

The eukcc_wgs-holobiont/workfiles/pplacer/placementReduced.jplace exists and is not empty. I tried re-running but got the same error.

[openpaul]

Hello, could you provide me with the eukcc_wgs-holobiont/workfiles/pplacer/placementReduced.jplace file (If possible all files in the pplacer folder would be perfect) ? It might be that no placement has enough support and is thus discarded. I might need to adjust the parameters and/or make the error message more meaningful.

[sarah872]

pplacer.zip

[openpaul]

So your placements had a posterior probability less than 0.4 (default), thus EukCC assumed they are not good enough to consider this placement valid. You can however change this threshold using the --minPlacementLikelyhood parameter. You could set it to 0 to deactivate this filtering. I would think thats also a valid parameter setting, although less conservative.

[openpaul]

Let me know if adjusting the parameter works for you. Might be worth writing about it in the documentation then.

[sarah872]

Yes, thank you, setting it to 0 worked!