I am in the process of assembling Illumina HiSeq transcripts from 3
different species into reference transcriptome databases. Neither
species has a reference genome. The first transcript set has approx
60,000 contigs. I am thinking about using Blast2GOpro for this work.
Based on experiences, what do you think the time required for
annotation will be? I understand there are a lot of variables that
can affect the speed, but a best guess would be appreciated. I have a
8core, 64bit CPU with 50G of RAM if that helps.
Thanks,
Steve
Steve, I would not recommend B2G unless you have already run Trinity. Also you need about 4x more computing power Bob ________________________________________ From: blast2go@googlegroups.com [blast2go@googlegroups.com] On Behalf Of Wiseman [steve.wise...@usask.ca] Sent: Monday, January 23, 2012 5:04 PM To: Blast2GO Subject: Blast2GO GoogleGroup Blast2go PRO
I am in the process of assembling Illumina HiSeq transcripts from 3 different species into reference transcriptome databases. Neither species has a reference genome. The first transcript set has approx 60,000 contigs. I am thinking about using Blast2GOpro for this work. Based on experiences, what do you think the time required for annotation will be? I understand there are a lot of variables that can affect the speed, but a best guess would be appreciated. I have a 8core, 64bit CPU with 50G of RAM if that helps. Thanks, Steve
-- You received this message because you are subscribed to the Google Groups "Blast2GO" group. To post to this group, send email to blast2go@googlegroups.com. To unsubscribe from this group, send email to blast2go+unsubscribe@googlegroups.com. For more options, visit this group at http://groups.google.com/group/blast2go?hl=en.
Hi Bob,
Thanks for replying. I have used CLC genomics workbench for assembly
of the Illumina data into Scaffolds/contigs. So, even then you say I
need 4x more computing power for the Blast2GOpro?
Steve
On Jan 23, 4:13 pm, Robert Chapman <Chapm...@dnr.sc.gov> wrote:
> Steve,
> I would not recommend B2G unless you have already run Trinity. Also you need about 4x more computing power
> Bob
> ________________________________________
> From: blast2go@googlegroups.com [blast2go@googlegroups.com] On Behalf Of Wiseman [steve.wise...@usask.ca]
> Sent: Monday, January 23, 2012 5:04 PM
> To: Blast2GO
> Subject: Blast2GO GoogleGroup Blast2go PRO
> I am in the process of assembling Illumina HiSeq transcripts from 3
> different species into reference transcriptome databases. Neither
> species has a reference genome. The first transcript set has approx
> 60,000 contigs. I am thinking about using Blast2GOpro for this work.
> Based on experiences, what do you think the time required for
> annotation will be? I understand there are a lot of variables that
> can affect the speed, but a best guess would be appreciated. I have a
> 8core, 64bit CPU with 50G of RAM if that helps.
> Thanks,
> Steve
> --
> You received this message because you are subscribed to the Google Groups "Blast2GO" group.
> To post to this group, send email to blast2go@googlegroups.com.
> To unsubscribe from this group, send email to blast2go+unsubscribe@googlegroups.com.
> For more options, visit this group athttp://groups.google.com/group/blast2go?hl=en.
I am not sure where Bob is getting his recommendations from. First, B2Go does not care where you get your scaffolds/contigs from. I am part of a core service that processes a lot of samples from other people. While I have not used CLC with B2Go, I have used Newbler (454) and Trinity output. Second, B2Go takes much less computing power and much less memory than Trinity. What B2Go does take a lot of is a network access and the tolerance for lag time. This is because B2Go queries its database very slowly (although the Pro version may be faster). Perhaps what Bob is thinking of is the actual Blasting. This can take a lot of computer resources if you want to do Blasts quickly. Or you can rely on NCBI or another global resource. But B2Go itself? A boring single-threaded "low memory" (say 16 GB maximum.) Compare that to Trinity which can easily use 140 GB and multiple cores.
> Hi Bob, > Thanks for replying. I have used CLC genomics workbench for assembly > of the Illumina data into Scaffolds/contigs. So, even then you say I > need 4x more computing power for the Blast2GOpro? > Steve
> On Jan 23, 4:13 pm, Robert Chapman <Chapm...@dnr.sc.gov> wrote: >> Steve, >> I would not recommend B2G unless you have already run Trinity. Also you need about 4x more computing power >> Bob >> ________________________________________ >> From: blast2go@googlegroups.com [blast2go@googlegroups.com] On Behalf Of Wiseman [steve.wise...@usask.ca] >> Sent: Monday, January 23, 2012 5:04 PM >> To: Blast2GO >> Subject: Blast2GO GoogleGroup Blast2go PRO
>> I am in the process of assembling Illumina HiSeq transcripts from 3 >> different species into reference transcriptome databases. Neither >> species has a reference genome. The first transcript set has approx >> 60,000 contigs. I am thinking about using Blast2GOpro for this work. >> Based on experiences, what do you think the time required for >> annotation will be? I understand there are a lot of variables that >> can affect the speed, but a best guess would be appreciated. I have a >> 8core, 64bit CPU with 50G of RAM if that helps. >> Thanks, >> Steve
>> -- >> You received this message because you are subscribed to the Google Groups "Blast2GO" group. >> To post to this group, send email to blast2go@googlegroups.com. >> To unsubscribe from this group, send email to blast2go+unsubscribe@googlegroups.com. >> For more options, visit this group athttp://groups.google.com/group/blast2go?hl=en.
> -- > You received this message because you are subscribed to the Google Groups "Blast2GO" group. > To post to this group, send email to blast2go@googlegroups.com. > To unsubscribe from this group, send email to blast2go+unsubscribe@googlegroups.com. > For more options, visit this group at http://groups.google.com/group/blast2go?hl=en.
Hi all! We do use clc data with b2g. The b2g input data can be whatever you want, the resources required for running your annotation is much less than to run blast or reads assembly. The pro version will speed up the mysql queries. In our lab we run b2g on a 55 k contigs dataset in about 2 weeks, maybe due to network issues. With locally installed database this was cut in half. Marcelo
Marcelo,
Thanks for the information. When you say it takes two weeks to run 55k
contigs, is that including Blast analysis? And are you using the Pro
version? I have started to Blast identify my 60000 contigs in CLC
workbench using a local database. Based on the first 24h it will take
about 20days to complete the analysis.
Steve
On Jan 24, 4:42 am, Marcelo Brandão <brandao.marc...@gmail.com> wrote:
> Hi all!
> We do use clc data with b2g. The b2g input data can be whatever you want,
> the resources required for running your annotation is much less than to run
> blast or reads assembly. The pro version will speed up the mysql queries.
> In our lab we run b2g on a 55 k contigs dataset in about 2 weeks, maybe due
> to network issues. With locally installed database this was cut in half.
> Marcelo
Hi Steve We use MPIblast, local interpro scan and the regular Blast2go accessing local database. You can use use the divide and conquer approach, divide your blast in several machines then import your xml's, the same could be done with interpro via b2g. This will speed up, a lot, your analyzes. This approach will demand a good computer to merge all your Xml's files. Cheers, Marcelo Em 24/01/2012 23:26, "Wiseman" <steve.wise...@usask.ca> escreveu:
> Marcelo, > Thanks for the information. When you say it takes two weeks to run 55k > contigs, is that including Blast analysis? And are you using the Pro > version? I have started to Blast identify my 60000 contigs in CLC > workbench using a local database. Based on the first 24h it will take > about 20days to complete the analysis. > Steve
> On Jan 24, 4:42 am, Marcelo Brandão <brandao.marc...@gmail.com> wrote: > > Hi all! > > We do use clc data with b2g. The b2g input data can be whatever you want, > > the resources required for running your annotation is much less than to > run > > blast or reads assembly. The pro version will speed up the mysql queries. > > In our lab we run b2g on a 55 k contigs dataset in about 2 weeks, maybe > due > > to network issues. With locally installed database this was cut in half. > > Marcelo
> -- > You received this message because you are subscribed to the Google Groups > "Blast2GO" group. > To post to this group, send email to blast2go@googlegroups.com. > To unsubscribe from this group, send email to > blast2go+unsubscribe@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/blast2go?hl=en.
If you use MPIblast (or any other form of Blast that does divide-and-conquer) then I am surprised that it takes you two weeks to run 55K contigs. For that size of data I can usually get the blasts done within 6 hours and then B2Go annotation within 24 hours. This using local databases. And also not doing any interpro work -- which might add significantly to time.
> Hi Steve > We use MPIblast, local interpro scan and the regular Blast2go accessing local database. You can use use the divide and conquer approach, divide your blast in several machines then import your xml's, the same could be done with interpro via b2g. This will speed up, a lot, your analyzes. This approach will demand a good computer to merge all your Xml's files. > Cheers, > Marcelo
> Em 24/01/2012 23:26, "Wiseman" <steve.wise...@usask.ca> escreveu: > Marcelo, > Thanks for the information. When you say it takes two weeks to run 55k > contigs, is that including Blast analysis? And are you using the Pro > version? I have started to Blast identify my 60000 contigs in CLC > workbench using a local database. Based on the first 24h it will take > about 20days to complete the analysis. > Steve
> On Jan 24, 4:42 am, Marcelo Brandão <brandao.marc...@gmail.com> wrote: > > Hi all! > > We do use clc data with b2g. The b2g input data can be whatever you want, > > the resources required for running your annotation is much less than to run > > blast or reads assembly. The pro version will speed up the mysql queries. > > In our lab we run b2g on a 55 k contigs dataset in about 2 weeks, maybe due > > to network issues. With locally installed database this was cut in half. > > Marcelo
> -- > You received this message because you are subscribed to the Google Groups "Blast2GO" group. > To post to this group, send email to blast2go@googlegroups.com. > To unsubscribe from this group, send email to blast2go+unsubscribe@googlegroups.com. > For more options, visit this group at http://groups.google.com/group/blast2go?hl=en.
> -- > You received this message because you are subscribed to the Google Groups "Blast2GO" group. > To post to this group, send email to blast2go@googlegroups.com. > To unsubscribe from this group, send email to blast2go+unsubscribe@googlegroups.com. > For more options, visit this group at http://groups.google.com/group/blast2go?hl=en.
> If you use MPIblast (or any other form of Blast that does > divide-and-conquer) then I am surprised that it takes you two weeks to run > 55K contigs. For that size of data I can usually get the blasts done > within 6 hours and then B2Go annotation within 24 hours. This using local > databases. And also not doing any interpro work -- which might add > significantly to time.
> On Jan 25, 2012, at 4:51 AM, Marcelo Brandão wrote:
> Hi Steve > We use MPIblast, local interpro scan and the regular Blast2go accessing > local database. You can use use the divide and conquer approach, divide your > blast in several machines then import your xml's, the same could be done > with interpro via b2g. This will speed up, a lot, your analyzes. This > approach will demand a good computer to merge all your Xml's files. > Cheers, > Marcelo
> Em 24/01/2012 23:26, "Wiseman" <steve.wise...@usask.ca> escreveu:
>> Marcelo, >> Thanks for the information. When you say it takes two weeks to run 55k >> contigs, is that including Blast analysis? And are you using the Pro >> version? I have started to Blast identify my 60000 contigs in CLC >> workbench using a local database. Based on the first 24h it will take >> about 20days to complete the analysis. >> Steve
>> On Jan 24, 4:42 am, Marcelo Brandão <brandao.marc...@gmail.com> wrote: >> > Hi all! >> > We do use clc data with b2g. The b2g input data can be whatever you >> > want, >> > the resources required for running your annotation is much less than to >> > run >> > blast or reads assembly. The pro version will speed up the mysql >> > queries. >> > In our lab we run b2g on a 55 k contigs dataset in about 2 weeks, maybe >> > due >> > to network issues. With locally installed database this was cut in half. >> > Marcelo
>> -- >> You received this message because you are subscribed to the Google Groups >> "Blast2GO" group. >> To post to this group, send email to blast2go@googlegroups.com. >> To unsubscribe from this group, send email to >> blast2go+unsubscribe@googlegroups.com. >> For more options, visit this group at >> http://groups.google.com/group/blast2go?hl=en.
> -- > You received this message because you are subscribed to the Google Groups > "Blast2GO" group. > To post to this group, send email to blast2go@googlegroups.com. > To unsubscribe from this group, send email to > blast2go+unsubscribe@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/blast2go?hl=en.
> -- > You received this message because you are subscribed to the Google Groups > "Blast2GO" group. > To post to this group, send email to blast2go@googlegroups.com. > To unsubscribe from this group, send email to > blast2go+unsubscribe@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/blast2go?hl=en.
-- Marcelo Mendes Brandão Postdoc fellow Young Investigator on Bioinformatics Laboratório de Biologia Molecular de Plantas - ESALQ/USP Website: http://bioinfo.esalq.usp.br AtPIN: http://bioinfo.esalq.usp.br/atpin SKYPE: mmbrand Tel: (+55) 19 3429 4442
