missing TRUE/FALSE error

[ptrebert]

Hi David,

this one is new and quite unexpected - did you change something in the code?

Environment: R 3.5.1 and StrandPhaseR_0.99 (from dev branch)

Loading VCF file ... 158.06s
Using 64 CPUs
Phasing chromosomes ...Error in { : task 4 failed - "cluster1
Error in if (X == Y & X != "" & Y != "") {: missing value where TRUE/FALSE needed
"
Calls: strandPhaseR -> %dopar% -> <Anonymous>
Execution halted

Best,
Peter

[ptrebert]

This error persists but only for the 732 individual...

Loading VCF file ... 134.09s
Using 24 CPUs
Phasing chromosomes ...Error in { : task 11 failed - "cluster19
Error in if (X == Y & X != "" & Y != "") {: missing value where TRUE/FALSE needed
"
Calls: strandPhaseR -> %dopar% -> <Anonymous>
Execution halted

I suspect there is some other error upstream of StrandPhaseR that is not handled properly (I assume it's an htslib problem...)

[ptrebert]

For future reference: test output created under [scratch] /users/pebert/testing/732_strandphaser/test

Observations:

• VCF for cluster19 in folder VCFfiles/ is empty
• there is no cluster19_reads.RData in the data/ folder (likely the cause for the above, I presume)
• the input alignment files do contain proper alignments for cluster19
• the input variant calls have called variants on cluster19

As far as I can tell, the input for the StrandPhaseR analysis is "as good" as for the other samples, so no clue what happens here...

[daewoooo]

Great let's resolve this after submission so don't mess with the results right now. Should be easy to fix. D.

…

On Mon, Nov 18, 2019 at 6:46 AM Peter Ebert ***@***.***> wrote: For future reference: test output created under [scratch] /users/pebert/testing/732_strandphaser/test Observations: - VCF for cluster19 in folder VCFfiles/ is empty - there is no cluster19_reads.RData in the data/ folder (likely the cause for the above, I presume) - the input alignment files do contain proper alignments for cluster19 - the input variant calls have called variants on cluster19 As far as I can tell, the input for the StrandPhaseR analysis is "as good" as for the other samples, so no clue what happens here... — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#4?email_source=notifications&email_token=ABXFIOLZLD4XZW7YZKT7GDDQUKTFXA5CNFSM4I6BG74KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEEKVMLQ#issuecomment-555046446>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABXFIOIQ4SYVGKQVBJK77ELQUKTFXANCNFSM4I6BG74A> .