[arne.muller]

Hello,

you mean exporting the current view into a graphics (e.g. png) file? I've had some relatively long response times during the export, but not as long as you report. I can imagine that the time needed for export is proportional to the number of elements in the current view. Maybe just take a screen shot for instead (not nice but often that's enough)?

Arne

[The_Roads]

No sorry I meant turning on alignment info like coverage maps, non-specific reads etc. it takes a very long time for CLC to generate the graph in the top frame. likewise once the graph is there it takes ages to export a csv file of the graph. i'd like to know if anyone else has this problem or whether it might be something funky with my workstation (win6x 1x quad xeon 32Gb)

Thanks

[smprince18]

disclaimer I work at CLC bio

The_Roads

I was wondering if you have tried to contact [email protected] yet? Or you can reach us at (617)-444-8765. It could be a result that your VMoptions where not adjust correctly by the installer. If you go to the directory for the CLCGenomicsWB3 and show hidden, then you will see the .vmoption file, open this in notepad, there will be a line that looks like the following -Xmx####, where # = the number of mb allocated to the application (an example Xmx1024, means you have 1 gb of RAM allocated to the application)

The second possibility you downloaded the 32 bit version of the application and not the 64 bit, this would result in very slow response time since, a 32 bit application can only request 2gb of ram)

Again if you would like some help with this contact our support team or myself directly.

Shawn M Prince

[The_Roads]

Hi Shawn,

Thanks I'll be in touch.

We have the 64 bit version and we've already tweaked vmoptions. Everything to relating to assembling, SNP detection etc. that requires 64 bit computing is working fine. It is just anything that alters the GUI or exports graphics/text that locks the workstations on large assemblies.

The_Roads

[polsum]

Hi, I have been testing the trial version of CLC workbench and I encountered two issues.

1. I BLASTed a set of illumina generated short reads (17-33, after trimming the 3'adapters) to mouse RefSeq database through stand alone BLAST program with stringent parameters and I found, say 2500 reads matching to mouse mRNAs. When I aligned all those 2500 reads to the same RefSeq database by using CLC reference assembly, only half of them are aligning to the reference. I tried all different options available changing the gap penalties, global alignment, scores etc...but never all the reads aligned to the reference. I think there should be more options here.

2. I used BLAST feature in CLC bench and when I view the blast output parsed results, I dont see all the columns in the overview table. For example I dont see strand orientation titled column in the overview. However, I see it in the individual blast mapping, but it is useless for me because I need to count the total number of minus and plus mappings of the total number of mappings. This is a serious limitation for me.

[The_Roads]

As an update for future readers, received some excellent help from CLC and it appears the delays we experienced were due to the way we assembled our contigs. Using conventional assembly parameters graphics now render in seconds to minutes (CLC3.6.5).

[smprince18]

Disclaimer I work for CLC bio

Polsum,

Once you Blast your sequences within CLC WB, you will be shown a graphic view of the results. If you look in the lower left hand corner of the working are you will see a table view. Once this is open you will see a default group of columns, Please note in the right hand side panel you can toggle on and off the columns you want to see.. You will be able to look at the direction of the results. Also if you are using this to reference map your reads we graphically show orientation (red = reverse read, green forward) Also the count of forward and reverse reads can be found in the contig report. Please let me know if this straightens anything up for you. If you would like I can be contacted at the CLC Boston office 617-444-8765.

Shawn

Disclaimer I work for CLC bio

[smprince18]

The Roads,

Glad to hear that your are enjoying your experience with CLC Genomics WB. Let me know if you need anything.

Shawn M Prince

Disclaimer I work at CLC bio

[johnny]

Hey there,

one more newbie here
I have a probably easy to answer question but can't find the solution on my own....
How can I see the corresponding protein sequence to a nucleotide sequence ? In more detail, after a reference assembly, I would like to click through the variations with the "Find Conflict" button and instantly see if the protein sequence is affected as well.

Thanks for your help!

[The_Roads]

Hi Johnny, I assume you are using CLCGWB. if so the conflict table is not the place to look. you should run a snp detection. if you have an annotated ref seq then the table will present you with all the amino acid changes.

[johnny]

Yes, this is what I did before, but I also had numerous "conflicts" in the consensus sequence but no hit in the SNP detection. After changing the sensitivity I now found these variation as well. So thanks for bringing me onto the right track again.
What do you think is the minimum coverage to call a SNP ?

[The_Roads]

Conflicts and SNPs are not really the same thing. In CLC a conflict is any variation from the ref seq or consensus. A SNP is a conflict that has passed quality and position requirements.
The minimum coverage for a SNP is debated a lot and obviously depends on your coverage and the type of project you are doing. For absolute reads, 8 reads that pass SNP calling is often used but i dont think there is really a gold standard, % all depends on your coverage. I would be very careful altering SNP quality or position criteria too much, you could easily end up with junk. CLC has excellent help on how SNPs are called. Vector and quality trimming and removal of broken reads should also improve the quality of your assembly before SNP detection. Gapped and ungapped assemblies can also give different SNP calls, particularly when there are conflicts between ref seqs and your consensus seqs. What type of project are you doing?

[smprince18]

************************disclaimer I work for CLC bio ***********************

Johnny,
The Roads gave you a good answer to run the SNP detection, b/c you will have set significant values for the SNP. The conflict table will reflect anywhere you see a difference, you could have 100x coverage 99 a 1 c and that will be a conflict, but may not be included in the SNP table (depends on your sig values). The snp table will also reflect any AA change within annotated regions.
If however you would still like to use the conflict table you can go to the right hand side panel, go to the heading Nucleotide info and choose to show your translation, you can pick any and all frames or tell it to translate ORF and CDS regions. If you have more questions feel free to contact me (shawn prince) at [email protected]

[johnny]

If however you would still like to use the conflict table you can go to the right hand side panel, go to the heading Nucleotide info and choose to show your translation, you can pick any and all frames or tell it to translate ORF and CDS regions. If you have more questions feel free to contact me (shawn prince) at [email protected]

This is what I was actually looking for in the beginning. But I think there are more questions rising up while digging deeper into that subject. So I'll probaly take your offer and write an email instead of bothering the community .
Thank you both for the quick reply!

[didymos]

Hi,

I am wondering how big amount of RAM you need to assembly long contigs from human genome sequencing with short Illumina reads. On the CLC web page is writen that you need much less then in case of SOAP which was used in assembly of Panda genome. To get good coverage you need about 150-200Gb of reads (>50x coverage). In Panda genome project they used supercomputer with 512GB RAM....