Human Genome Project Information Archive
1990–2003

JGI Sequencing: Moving Toward a Consensus Strategy

Chris Martin (LBNL) and Jane Lamerdin (LLNL) outlined progress toward forming efficient collaborations among the member laboratories and establishing a facility capable of sequencing 100 Mb per year. They observed that this ramp-up will require increasing the scale of operations (hiring more people) and improving the processes in terms of cost, throughput, and quality.

At LBNL, emphasis is on the up-front shotgun phase, using double-end plasmid subclone sequencing and moving toward increasingly automated finishing processes. The primary templates are BACs for chromosome 5, with gaps finished via transposons. Martin reported working closely with LANL to establish sequence-annotation and -submission processes. LBNL ships completed clones to LANL, where restriction-digest verification required for JGI quality is performed.

At LLNL, sequencing targets are chosen from regions of the high-resolution chromosome 19 physical map, which offers 42 Mb (fivefold to sixfold coverage) of sequence-ready cosmid and BAC clones, Lamerdin said. About 90% of genomic targets are 1 Mb, and the remaining 10% are regions of at least 100 kb that contain genes of interest. All regions are flanked by publicly available markers. JGI targets can be found via the National Center for Biotechnical Information (NCBI) Web site [http://www.ncbi.nlm.nih.gov/].

Investigators follow a standard shotgun strategy using M13 and plasmid vectors and a mixture of sequencing chemistries, with automated base calling and clone sequence assembly by Phred and Phrap. Sequences are edited using the CONSED package, and gaps are closed with a variety of techniques, including PCR and an in vitro transposon technology. Each sequence is verified relative to three independent restriction digests, and sequence analysis and annotation are done using local BLAST and GRAIL coding-prediction tools. Results are parsed manually and submitted through SEQUIN to NCBI.

Lamerdin said the member laboratories are instituting quantitative quality standards for JGI, which will be adopting these processes for all generated sequence. (For details, see Bermuda-Quality Sequence) This standard will enable data to be compared within JGI and with sequence generated by NIH centers. David Nelson is monitoring data-quality processes and helping to implement them. All data are available on the Web site, and submitted data are in GenBank. All sites have implemented a Web-accessible list of accessions, submit lengths, and Phrap-quality statistics. Because investigators are focusing on genomic continuity, they also have a "unique length" category indicating sequence new to the database.

Martin and Lamerdin also described some successful methods for increasing sequence-data quality and throughput and for decreasing costs. At LBNL, significant improvements in automation were made with commercial Tecan and custom Prep Track robots. Other changes include switching to a Qiagen real prep for DNA preparation before the shotgun phase, using BigDye Terminators in sequencing reactions, and moving to 64-lane gels. Changes at LLNL instituted by Paula McCready include incremental chemistry improvements (i.e., use of the polymerase Omnivase from Promega), in-house reagent quality control for all buffers, and production and finishing operations. In collaboration with JGI, the Whitehead Institute is developing front-end automation and sequencing on chromosome 19.

The electronic form of the newsletter may be cited in the following style:
Human Genome Program, U.S. Department of Energy, Human Genome News (v9n3).