SEA-PHAGES | BLAST in DNAM

Link to this post \| posted 29 Feb, 2016 01:59
kwhutch	I'm in the 70+ million seconds for a single gene BLAST. I'm supposedly retired so I guess that means I can wait that long. I will say that having this Forum and getting pretty quick information that the problem was someplace else and not that my computer was going through a slow burn is comforting. Keith

Link to this post \| posted 29 Feb, 2016 19:34
debbie	Hi all! I was just able to blast a gene (and repeated by Dan) in DNA Master. The Integer Overflow error is gone and my posted wait time was 31 seconds, with the blast data returned in about 1 1/2 minutes. I just started to Blast an entire genome and will keep you posted as to how that goes! We are back in business!

Link to this post \| posted 29 Feb, 2016 19:48
DanRussell	Debbie Jacobs-Sera Hi all! I was just able to blast a gene (and repeated by Dan) in DNA Master. The Integer Overflow error is gone and my posted wait time was 31 seconds, with the blast data returned in about 1 1/2 minutes. I just started to Blast an entire genome and will keep you posted as to how that goes! We are back in business! Hopefully this is fixed, everyone! Nice job by Debbie bugging NCBI into cooperation. Let us know if you continue to have problems. You shouldn't need to update anything to have success, just try again. –Dan

Link to this post | posted 29 Feb, 2016 19:48

DanRussell

Debbie Jacobs-Sera
Hi all! I was just able to blast a gene (and repeated by Dan) in DNA Master. The Integer Overflow error is gone and my posted wait time was 31 seconds, with the blast data returned in about 1 1/2 minutes. I just started to Blast an entire genome and will keep you posted as to how that goes! We are back in business!

Hopefully this is fixed, everyone! Nice job by Debbie bugging NCBI into cooperation.

Let us know if you continue to have problems. You shouldn't need to update anything to have success, just try again.
–Dan

Link to this post \| posted 29 Feb, 2016 21:35
saleadon	Thanks Debbie.

Link to this post \| posted 29 Feb, 2016 23:43
debbie	Success! I blasted an entire phage genome (52KB) and finished the blast. The blast log details that the blast finished in about an hour, but it could not retrieve all of the blast data for about 3 additional hours. However, the blast did finish and it is looking good so far!

Link to this post \| posted 01 Mar, 2016 18:40
kwhutch	Debbie Jacobs-Sera Success! I blasted an entire phage genome (52KB) and finished the blast. The blast log details that the blast finished in about an hour, but it could not retrieve all of the blast data for about 3 additional hours. However, the blast did finish and it is looking good so far! Debbie, Thanks for your work with NCBI to get this running again. I made a run last night with a similar time frame from start to finish. I started mine after 9pm because when I start a run during the peak hours the time from start to finish can double or more. So I will just raise this question and maybe some clever person will know how to tweak DNA Master in some way to address my question. The question is why does DNA Master take so long? Today at midday I took a FASTA file of the peptide sequences for the 99 ORFs that BLASTed last night and dumped it into blastp at NCBI and BLASTed against the full nr database. Last night with DNA Master in off peak hours it took about 3 hours. Using the web interface approach I had the results page back from NCBI in about 9 minutes. It took another couple minutes to download the results to my computer. I could have chosen any of several formats for the download. For simplicity I chose a text file. One can specify up to a list of 100 hits and 50 alignments for each ORF in the file. For space I just did 10 each. Still the whole process was done in under 15 minutes. All that remains is how to get the data into DNA Master. Again, I don't have an answer but I assume there must be one. In the meantime I will BLAST at night or weekends. Keith

Link to this post | posted 01 Mar, 2016 18:40

kwhutch

Debbie Jacobs-Sera
Success! I blasted an entire phage genome (52KB) and finished the blast. The blast log details that the blast finished in about an hour, but it could not retrieve all of the blast data for about 3 additional hours. However, the blast did finish and it is looking good so far!

Debbie, Thanks for your work with NCBI to get this running again. I made a run last night with a similar time frame from start to finish. I started mine after 9pm because when I start a run during the peak hours the time from start to finish can double or more.

So I will just raise this question and maybe some clever person will know how to tweak DNA Master in some way to address my question.

The question is why does DNA Master take so long? Today at midday I took a FASTA file of the peptide sequences for the 99 ORFs that BLASTed last night and dumped it into blastp at NCBI and BLASTed against the full nr database. Last night with DNA Master in off peak hours it took about 3 hours. Using the web interface approach I had the results page back from NCBI in about 9 minutes. It took another couple minutes to download the results to my computer. I could have chosen any of several formats for the download. For simplicity I chose a text file. One can specify up to a list of 100 hits and 50 alignments for each ORF in the file. For space I just did 10 each. Still the whole process was done in under 15 minutes. All that remains is how to get the data into DNA Master.

Again, I don't have an answer but I assume there must be one. In the meantime I will BLAST at night or weekends.

Keith

Link to this post \| posted 02 Mar, 2016 18:30
debbie	Keith, I asked Dr. Lawrence about this and here is his response: "Yes, DNA Master will wait until the BLAST server send replies; the delays are due to the failures of the BLAST server to return results in a timely fashion. Outside of abandoning searches and resubmitting the data - which they explicitly tell you not to do as it simply overloads the servers - there it nothing one can do. Their policy does ignore the possibility that the server may never return the results, or return them after an inordinate amount of time." Not very satisfying, but may help us to understand. I blasted a genome yesterday that produced the usual errors and took 8 hours to complete, but it did finish.

Link to this post | posted 02 Mar, 2016 18:30

debbie

Keith,
I asked Dr. Lawrence about this and here is his response:
"Yes, DNA Master will wait until the BLAST server send replies; the delays are due to the failures of the BLAST server to return results in a timely fashion. Outside of abandoning searches and resubmitting the data - which they explicitly tell you not to do as it simply overloads the servers - there it nothing one can do. Their policy does ignore the possibility that the server may never return the results, or return them after an inordinate amount of time."
Not very satisfying, but may help us to understand. I blasted a genome yesterday that produced the usual errors and took 8 hours to complete, but it did finish.

Recent Activity

BLAST in DNAM