Transeq bug

[anilthanki]

EMBOSS TranSeq has a bug/feature which removes anything before a colon : from sequence ids (i.e. the first "word" after > in each FASTA header line).

This results in empty cluster_cds files and failed workflow.

So if input sequences contains : in sequence identifiers, they need to be removed or another copy of CDS dataset needs to be prepared to avoid this error.

Thanks

Anil

[nsoranzo]

Example:
Input:

> Before colon: after colon
accaGTTACCCTCATCATCTTAGCTGATAGCCAGCCAGCCACCACAGGCAtgagtca
> Before_colon:after_colon
gtttgccatcttttgctgctctagggaatccagcagctgtcaccatgtaaacaagcccagg

Output of transeq -sequence 'test.fasta' -outseq 'protein.fasta' -frame 1 -table 0 -trim no -osformat2 fasta -auto:

>Before_1 colon: after colon
...
>after_colon_1
...

[anilthanki]

Extended test

>test:A
ACAGTCATCGAATCCGACTAC
>test : B
ACGACTAGCATCAGCACTA
>test :C
ACGACTAGCATCAGCACTA
>test: D
ACGACTAGCATCAGCACTA

Output

>A_1
TVIESDY
>test_1 : B
TTSISTX
>test_1 :C
TTSISTX
>test_1 D
TTSISTX