EdgeBio is committed to testing and validating the latest technologies from Ion Torrent to provide the latest sequencing services to our clients as soon as possible. Recently we successfully performed some pilot runs using the new long read chemistry. Although we haven't yet achieved the high loading we used to achieve with short read chemistry, we are pleased with the run quality and throughput from our pilot runs and want to share some of the data with the community.

Many of the sequencing runs released so far from Ion Torrent (with the exception of two new HuRef 318 chip datasets available here) have been resequencing of DH10B. While E. coli is a good organism for validation and testing, we see human sequencing as the main Ion Torrent application for us in the future. For this reason we typically run both DH10B and hg19 pilot runs when testing new technology. The run we are releasing today is shotgun resequencing of Craig Venter HuRef DNA. While the mean coverage depth and percent coverage are quite low and aren't really useful for variant calling, the high throughput of AQ20 data from this run shows the Ion's potential for targeted human resequencing.

GenoLogics today announced that EdgeBio, a leading sequencing services provider, has deployed the GenoLogics Laboratory Information Management System (LIMS), bringing an unprecedented level of flexibility and adaptability to its next-generation sequencing (NGS) core laboratory. The GenoLogics LIMS replaces EdgeBio’s existing commercial LIMS software.

Last Wednesday I had the privilege of giving a talk at a local Ion Torrent user group meeting.

With the announcement of the new RecognitION program EdgeBio is proud to release our latest run of DH10B resequencing data to the community.  Our latest run, using OneTouch prep, yielded 414.31MB (350.93MB Q20) of sequencing data.  We're really excited to push the platform to its limits and hope that releasing data to the community will help others do the same.

You can view the run report in any web browser by clicking on the link below using the username 'edgecommunity' and leaving the password blank:

Variant calling, the detection of SNPs and INDELs, plays a particularly important role in Ion Torrent data due to its propensity for homopolymer errors.  It’s particularly challenging to sort the insertions and deletions that occur through sequencing errors from true differences from the reference sequence.  A variant calling plugin is included in the Ion Torrent analysis pipeline that assists in the identification of SNPs and INDELs.  Utilizing SAMtools the plugin produces a variant sample report using settings adjusted to match the error model in Ion Torrent data.  Using recently sequenced E. coli DH10B data from a 316 chip and an artificially mutated E. coli genome the variant analysis plugin settings were compared to other samtools settings to try to find settings that produce the most true variants while avoiding false positives.  This mutated genome and a genome comprised of only homopolymer errors were also used to compare Illumina’s MiSeq technology to Ion Torrent in terms of variant analysis.