Hi Jim,

Unfortunately, the nodes in the cluster I'll probably be using only have 2GB 
per CPU, so I'll be better off if I can partition the task into two 2GB chunks.

So, the situation is that I don't have a dedicated server.  I have users who 
might have (at most) 20 pairs of primers that they will want to validate at 
once (i.e. "do these primers target my gene / region of interest?"  "What's the 
size of the product they'll produce?" "Is there anything else they target?").  
It's for a research lab at Mayo Clinic, and it MIGHT get used, at most, a 
couple of times a day.

My reading is that while we could send our users to the UCSC In-Silico PCR Web 
page, we can NOT access it from our own program.  Correct?

Am I going to be spending minutes launching isPcr, and then seconds actually 
getting results?  Is it possible to pre-compute the indexes, and then just load 
them in as necessary?

Thank you,

Greg

----- Original Message -----
From: "Jim Kent" <[email protected]>
To: "Gregory Dougherty" <[email protected]>
Cc: [email protected]
Sent: Tuesday, July 20, 2010 7:45:52 PM GMT -06:00 US/Canada Central
Subject: Re: [Genome] Finding where Primers hit

Hi - if you are doing a lot of primers at once usually you want to use isPcr.   
If you want to have a server set up that will quickly align a few primers at a 
time use gfPcr.  The memory these days is generally not a problem.  I think 4 
gig is enough for the whole genome for isPcr.


On Jul 20, 2010, at 11:56 PM, Gregory Dougherty wrote:

> Thank you Hiram and Jim for pointing me in the right direction.
> 
> Looking at the code, for my situation I have two workable choices: gfPcr or 
> isPcr.
> 
> Questions:
> 1: Is gfPcr actually a workable choice?  I.E. will it let me search for hits 
> from 18 base primers, and will it be as successful as isPcr at finding hits?
> 
> 2: The ReadMe says that isPcr "builds its own index".  Is it creating it's 
> own index, or loaded an already calculated one from files?
> 
> 3; For isPcr, roughly how much is "a lot of memory"?  In particular, roughly 
> how much space does it take from Chr1, and roughly how much does it take for 
> the whole human genome?
> 
> Thanks again,
> 
> Greg
> 
> ----- Original Message -----
> From: "Jim Kent" <[email protected]>
> To: "Gregory Dougherty" <[email protected]>
> Cc: [email protected]
> Sent: Tuesday, July 20, 2010 1:52:22 PM GMT -06:00 US/Canada Central
> Subject: Re: [Genome] Finding where Primers hit
> 
> Please use In Silico PCR, which was designed just for this purpose.
> 
> On Jul 20, 2010, at 4:51 PM, Gregory Dougherty wrote:
> 
>> Hi all,
>> 
>> I hope this isn't a FAQ, I couldn't find any way to search the archives.
>> 
>> I am writing a program that takes pairs of PCR Primer sequences, and returns 
>> the products they will produce when used w/ human DNA.  My first thought was 
>> to use BLAT to figure out where the primers match the genome.  While this 
>> has worked well with test primers that are 20+ bases long, BLAT won't run on 
>> smaller Primers, and my users have Primers as small as 18 bases.
>> 
>> Is BLAT a reasonable tool to use for solving this problem?  If we download 
>> the source, change the 20 base limit to 18, and then run that BLAT, will 
>> that work?  Work only if the sequence is a perfect match?  Fail miserably?
>> 
>> May be OT: Can I get Blast to tell me WHERE the primer hit in the item it 
>> hit in?  "Homo sapiens chromosome 17 genomic contig, GRCh37 reference 
>> primary assembly, Length=21169982" really isn't a very useful hit report.
>> 
>> Thanks in advance,
>> 
>> Greg
>> _______________________________________________
>> Genome maillist  -  [email protected]
>> https://lists.soe.ucsc.edu/mailman/listinfo/genome
> 
> _______________________________________________
> Genome maillist  -  [email protected]
> https://lists.soe.ucsc.edu/mailman/listinfo/genome

_______________________________________________
Genome maillist  -  [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome

Reply via email to