Just got back from the Complete Genomics User Group Meeting in San Francisco. Hopefully we'll start seeing some discussion of Complete data here on SEQanswers in the near future.
I have yet to take a close look at our data back from Complete yet, but I now have a much stronger understanding of their bioinformatics and technical pipeline (you can get their biochemistry info from their white paper). Anyone else here get data from them or attend the meeting?
I think CGA tools is very powerful for small variant detection. It looked from the data presented (again, preliminary) that their variant detection accuracy is on par or a little better than what is generally seen with Illumina or SOLiD whole genome data.
Given the nature of their data, converting to BAM and doing alternative variant analysis may not be as powerful (this is something we users may have to try). This is because they do assembly over non-reference called positions in their data processing. It will be interesting to see how their variant detection performance compares to variants detected on other platforms and with other analysis platforms.
They are just starting to produce CNV and SV calls. These data I'm not sure about yet, but it is good to see them diving into it with both feet because these variants are at least as important as small variants from a whole genome perspective.
They are also heading towards in-house somatic variants analysis (comparing two genomes), which is pretty important for cancer genomics. It will be interested to see what is delivered there in the future. Right now, the CGAtools calldiff algorithm is used for this. I'd like to hear opinions on calldiff from people who've used it.
Here's a pic of their machines in action:
![](http://lh5.ggpht.com/_78W20ysICjQ/TLYl8jRITAI/AAAAAAAAAjI/7GisfG7LrS4/s512/2010-10-12_19-01-14_922.jpg)
It's apparently 80% humidity and hot in that room. Yes, it's a dark room. This is because it's the optimal environment for the DNBs (DNA nano-balls) and for the picoliter quantities of reagents used at each cycle. (Later on, we had a rave party in there when the sequencers were on their lunch break!)
Here's one of their slides:
![](http://lh5.ggpht.com/_78W20ysICjQ/TLYlulCYBSI/AAAAAAAAAig/W5pa4FjxFIM/s512/2010-10-12_18-52-17_411.jpg)
Complete Genomics may be a strong way to go in the near future for whole genome sequencing. I'd certainly consider it as an alternative to doing it in-house at this point, freeing up local machines for targeted pull-downs, exomes, non-human, RNAseq, ChIPseq, etc.
I look forward to hearing from other Complete Genomics users in the future.
I have yet to take a close look at our data back from Complete yet, but I now have a much stronger understanding of their bioinformatics and technical pipeline (you can get their biochemistry info from their white paper). Anyone else here get data from them or attend the meeting?
I think CGA tools is very powerful for small variant detection. It looked from the data presented (again, preliminary) that their variant detection accuracy is on par or a little better than what is generally seen with Illumina or SOLiD whole genome data.
Given the nature of their data, converting to BAM and doing alternative variant analysis may not be as powerful (this is something we users may have to try). This is because they do assembly over non-reference called positions in their data processing. It will be interesting to see how their variant detection performance compares to variants detected on other platforms and with other analysis platforms.
They are just starting to produce CNV and SV calls. These data I'm not sure about yet, but it is good to see them diving into it with both feet because these variants are at least as important as small variants from a whole genome perspective.
They are also heading towards in-house somatic variants analysis (comparing two genomes), which is pretty important for cancer genomics. It will be interested to see what is delivered there in the future. Right now, the CGAtools calldiff algorithm is used for this. I'd like to hear opinions on calldiff from people who've used it.
Here's a pic of their machines in action:
![](http://lh5.ggpht.com/_78W20ysICjQ/TLYl8jRITAI/AAAAAAAAAjI/7GisfG7LrS4/s512/2010-10-12_19-01-14_922.jpg)
It's apparently 80% humidity and hot in that room. Yes, it's a dark room. This is because it's the optimal environment for the DNBs (DNA nano-balls) and for the picoliter quantities of reagents used at each cycle. (Later on, we had a rave party in there when the sequencers were on their lunch break!)
Here's one of their slides:
![](http://lh5.ggpht.com/_78W20ysICjQ/TLYlulCYBSI/AAAAAAAAAig/W5pa4FjxFIM/s512/2010-10-12_18-52-17_411.jpg)
Complete Genomics may be a strong way to go in the near future for whole genome sequencing. I'd certainly consider it as an alternative to doing it in-house at this point, freeing up local machines for targeted pull-downs, exomes, non-human, RNAseq, ChIPseq, etc.
I look forward to hearing from other Complete Genomics users in the future.
Comment