DnoiSE does not complete: "KeyError: 'count'"

[SanniH]

Hi!

I tried to run the DnoiSE.py with my own fasta file of dereplicated, non-chimeric ESVs, but unfortunately I got an error after about 45min, and was hoping you might be able to help?

The error I get:

Traceback (most recent call last):
File "DnoisE/src/DnoisE.py", line 467, in
denoised_ratio_d[i]['sequence'].upper() + "\n")
KeyError: 'count'

I run DnoiSE on a remote server through a shell script, and my fasta is formatted as requested following your paper and the repo, although there was no mention of whether all bases must be as capitals so I have a mix.

$ head UTILA_DSE.fasta

uniq1;size=62441;
TTATTCTACATACCCTGCTAGTGCTTATTTATCAACTGATTTAATAATCTTTTCATTACATTTAGCCGGTGCTAGTTCTATATTGTCTTCAATAAATTTTATTATAACAGTTTTTATGTTGCCTATAAattcttctttttctttttttcaatatcctttatttatagtagctcaaattactgtttcttttttATTATTAATATCTTTACCTGTTTTAGCCGCTGCTATTACTATGTTACTTTTTGATCGTAATTTCAACACTTCttttttttCCAATTATTTGGGTGGTGATGCTCTTCTTTATCAACATTTATTT
uniq2;size=24836;
TTTGAGTAGTGTTCAAGCTCATTCAGGTCCTTCTGTGGATTTGGCTATTTTTAGCCTTCATTTGTCCGGGGCAGCATCTATTATGGGTTCGATTAATTTCATTACAACAATTATTAATATGCGACCGGGAGGAATGGGAATGCATCGTTTGCCGCTATTTGTATGGGCAGTTTTGCTAACCGCAATTCTATTGTTGCTTTCTCTTCCTGTTTTGGCTGGGGGTATTACTATGTTGTTGACTGACCGAAATTTTAACACTACCTTTTTTGATCCCGCTGGAGGAGGAGACCCTGTTCTTTATCAACACCTATTT
...

My file was generated using PEAR for PE merging, and VSEARCH for length filtering and dereplication, and the resulting fasta contains ~44K unique sequences, single line, with length ranging from 303-323 (COI Leray fragment). I attached the fasta I used here.

The command I used for running DnoiSE was this:
$ python3 DnoisE/src/DnoisE.py -i UTILA_DSE.fasta -o Utila -c 20

Any help would be appreciated as I am very keen to see how this compares to my previously generated data using dada2!

UTILA_DSE.zip

[adriantich]

Hi! Thanks for contacting me. Since many names refer to size/count/reads depending on the input format, I designed DnoisE to run whatever the name is. However, now I see that it was not done for the fasta_ouput line. I'll fix it now and tell you when it's done! I'll also modify to accept both upper and lower case letters and combined. Thank again! Adrià

…

On 16/4/21 13:58, SanniH wrote: Hi! I tried to run the DnoiSE.py with my own fasta file of dereplicated, non-chimeric ESVs, but unfortunately I got an error after about 45min, and was hoping you might be able to help? The error I get: Traceback (most recent call last): File "DnoisE/src/DnoisE.py", line 467, in denoised_ratio_d[i]['sequence'].upper() + "\n") KeyError: 'count' I run DnoiSE on a remote server through a shell script, and my fasta is formatted as requested following your paper and the repo, although there was no mention of whether all bases must be as capitals so I have a mix. $ head UTILA_DSE.fasta uniq1;size=62441; TTATTCTACATACCCTGCTAGTGCTTATTTATCAACTGATTTAATAATCTTTTCATTACATTTAGCCGGTGCTAGTTCTATATTGTCTTCAATAAATTTTATTATAACAGTTTTTATGTTGCCTATAAattcttctttttctttttttcaatatcctttatttatagtagctcaaattactgtttcttttttATTATTAATATCTTTACCTGTTTTAGCCGCTGCTATTACTATGTTACTTTTTGATCGTAATTTCAACACTTCttttttttCCAATTATTTGGGTGGTGATGCTCTTCTTTATCAACATTTATTT uniq2;size=24836; TTTGAGTAGTGTTCAAGCTCATTCAGGTCCTTCTGTGGATTTGGCTATTTTTAGCCTTCATTTGTCCGGGGCAGCATCTATTATGGGTTCGATTAATTTCATTACAACAATTATTAATATGCGACCGGGAGGAATGGGAATGCATCGTTTGCCGCTATTTGTATGGGCAGTTTTGCTAACCGCAATTCTATTGTTGCTTTCTCTTCCTGTTTTGGCTGGGGGTATTACTATGTTGTTGACTGACCGAAATTTTAACACTACCTTTTTTGATCCCGCTGGAGGAGGAGACCCTGTTCTTTATCAACACCTATTT ... My file was generated using PEAR for PE merging, and VSEARCH for length filtering and dereplication, and the resulting fasta contains ~44K unique sequences, single line, with length ranging from 303-323 (COI Leray fragment). I attached the fasta I used here. The command I used for running DnoiSE was this: $ python3 DnoisE/src/DnoisE.py -i UTILA_DSE.fasta -o Utila -c 20 Any help would be appreciated as I am very keen to see how this compares to my previously generated data using dada2! UTILA_DSE.zip <https://github.com/adriantich/DnoisE/files/6324683/UTILA_DSE.zip> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#8>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASBYEQHQOJB2CKTAADJDQGDTJAQ7JANCNFSM43BMWKDA>.

[adriantich]

Hi SanniH,
The problem is fixed.
you can update using git pull. Tell me if worked well please!
A.

[SanniH]

Hi Adria,

I'll put it to run now and report back if there's any more issues :)
Thanks for the quick response!

Sanni