|
|
|
Report of Interim Working Group Meeting
Miami Beach, Florida
September 17, 1998
Funding:
Japan: Takuji Sasaki reports that his group has received funding of US$10M for 10 years subject to review.
France: Michel Delseny reports that initial work to prepare sequence ready BACs will begin in his lab in Perpignan and in Alain Ghesquiere's and Jean Christophe Glaszmann's labs at ORSTOM/CIRAD in Montpellier, and that the work will move to Genoscope in Evry for sequencing.
Taiwan: Yue-ie Hsing reports that the National Science Council, Academia Sinica and the Education Ministry will support rice genome sequencing at 3 to 5 sites with 2 to 4 PIs in each site.
Korea: Moo Young Eun reports that they have applied the 1st 5-year(total 10-year) special budget for rice sequencing. from their government. For FY99 this would support 1.2 Mb of sequencing and new automated sequencers. The proposal passed several reviewing steps and is in final evaluation period. He is optimistic. In the meantime, they have been conducting pilot sequencing with their BAC clones, and will begin sequencing Nipponbare as soon as the requested libraries arrive.
Canada: Tom Bureau at McGill University has received a Provincial equipment grant to establish a sequencing facility from which he hopes to launch a Canadian rice genome sequencing program.
China: Guofan Hong reports that his group receives funds from the State Commission of Science and Technology, the Chinese Academy of Sciences and the Shanghai Municipal Government to support the current project of sequencing chromosome 4 of the rice genome at the rate set in the International Rice Genome Sequencing Areeement. The funding period is 5 years and may be renewed based on review.
US: Rod Wing (Clemon University) received funding from the Rockefeller Foundation to prepare BAC libraries. He has also received funding from Novartis to do BAC end sequencing and fingerprinting of the BACs. All of these results are being made publicly available. Machi Dilworth (NSF) reported that an interagency program announcement for sequencing rice, with USDA as the lead agency, would be published early in the 1999 fiscal year beginning October 1.
IRRI: While strictly not part of the Rice Genome Sequencing Program, it is important to note that Hei Leung has begun a functional genomics program based on deletion mutagenesis. He has used primarily fast neutrons and diepoxybutane to generate M3 populations with mutation rates of 1/1000 per locus and has obtained preliminary evidence supporting this work at the Xa21 locus conditioning bacterial blight resistance.
Progress:
Japanese Rice Genome Research Program:
Tomoya Baba reported that the RGP mapped 2000 ESTs on a YAC map of the rice genome previously constructed with 2275 DNA rice genome markers. They achieved 306 Mb of total coverage with 82% of chromosome 6 and 74% of chromosome 1 covered. Baba pointed out that particularly on chromosome 6, the gaps corresponded to regions of the map that lacked DNA markers.
The RGP constructed PAC libraries with the pCYPAC2 vector from Pieter de Jong. The vector employs the SacB gene for positive selection of inserts. They used partial Sau3AI digests to prepare their library which comprises 71,000 clones. The mean insert size is 112kb with 18% under 80 kb and 11% chloroplast contamination. The RGP is using a combination of restriction digests and hybridization or PCR of ESTs to establish contigs. They plan to obtain PAC end sequences using tail PCR. A second PAC library made with partial MboI digests of genomic DNA is now under construction.
Jo Messing:
Jo presented an overview of the International Agreement for rice genomic sequencing with specific emphasis on methodology. In particular, BAC end sequencing and shot gun sequencing. He illustrated his talk with assembled sequences from syntenous regions of sorgum and maize. He pointed out that it was important to have long, single pass reads from both ends of a random clone as distance is very important in assembling repeated sequences. The LTR regions of some retrotransposons that his group discovered exceeded 4kb, and these repeated sequences could not have been assembled without this protocol. In summarizing his talk, Jo showed how sequencing from individual BAC clones was both important and necessary.
Jo is collaborating with the Clemson group. He confirmed quality checks of the Nipponbare BAC library with BAC end-sequencing (see below). He also selected BAC clones generated by Rod Wing's group from japonic and indica with an orthologous probe from maize for a joint DNA sequencing project to compare sequence differences between the two subspecies.
Rod Wing:
Rod described the work of the Clemson Genomics Institute. A BAC library prepared with partial HindIII digest of Nipponbare DNA has produced a library of 36, 864 clones with an average insert size of 128.5 kb of which 8.2% have inserts less than 80kb. The library represents a greater than 10 fold coverage of the genome. 2.25% of the inserts appear to be chloroplast DNA and have been removed from the library. The library was probed with 12 RFLP single-copy markers and each probe produced from 2 to 21 hits. An EcoRI partial library is also being constructed.
The BAC libraries will be end sequenced with a goal of obtaining 148,000 ends. Currently 13,932 reads have been obtained with an average read length of 380 bp. These sequences are released monthly. 8432 sequences are their BLAST hits are currently posted on the Genome Institutes web pages, which also provide for searching by sequence.
Rod's colleague, Ralph Dean, spoke about finger printing the BACs. Their goal is to fingerprint and assemble 74,000 BACs. They are currently proving the technology on the rice blast fungal pathogen.
Dick McCombie:
Dick spoke about lessons learned from the Arabidopsis sequencing project with specific emphasis on the usefulness of a physical map based on BACs. He pointed out that the current cost of obtaining finished sequence is US$ 0.4 to 0.5 per base with the possibility that this might drop to US$ 0.3. Thus the cost of the rice sequencing project will be at least US$150M. The rice sequence will provide value in the form of rice improvement, the identity of almost all cereal genes, and a reference for larger cereal genomes based on comparative genomics. In order for this value to be realized, the sequence needs to be accurate, complete, and correlated with genetic markers and maps.
The group at Washington University headed by John McPherson and Marco Marra has undertaken a whole genome approach to fingerprinting Arabidopsis. Mapped markers or ESTs are used to anchor the physical map to the genetic and cytogenetic maps. The group has fingerprinted BACs from two libraries representing an 18-fold coverage of the genome. 17,000 clones have been assembled into 394 contigs. Dick pointed out that the automated assembly gives you bins (although they need to be checked) but does not give you precise order information without manual editing at the rate of 15 Mb per month for an experienced person. Currently, 75 contigs (comprising 13,000 clones) have been edited and cover 81 Mb.
Dick estimated that it would cost about US$ 2M to map the rice genome and that it would take an experienced team two years to complete the task. Dick pointed out that recent advances in mapping technology and strategies allow the task to be done quickly for a fraction of the former cost. In addition to greatly aiding sequencing, the map is likely to pay for itself as a tool for positional cloning prior to the completion of the sequence.
Proposal for a whole genome approach to rice sequencing:
Ian Bancroft and Mike Bevan submitted "A whole-genome approach to sequencing the rice genome" for discussion at this meeting. In their absence, the proposal was outlined by George Murphy.
While a number of people found the proposal attractive for larger genomes, there was little support for this proposal for rice. Basically, the question was do you want to discover the most genes in the shortest possible time or to you want to proceed most efficiently toward a complete reference sequence? The group assembled in Miami was decidedly of the second opinion. In this context, there were three areas that presented problems:
1) A number of people said that you needed sequence plus position for rice as a model organism and they didn't believe that the proposed outline would lead to map-based sequence.
2) If, as outlined in the proposal, most of the genes can be found in 20% of the DNA, it was not obvious how you would use this method to obtain the complete sequence.
3) The principal assumptions, that most of the genes can be discovered from a narrow range of buoyant density and that whole-genome shotgun reads can be assembled, are untested. Attendants suggested two preliminary experiments. The first was to prepare clones from the "gene-space DNA" as well as from the whole genome, and examine the frequency of homologous hits in sample single pass reads from both sets of clones. The second was to assemble the random single pass reads, but not the finishing reads, from the extensive contiguous sequence already obtained. For example, the 2 Mb that the European consortium has obtained from Arabidopsis would represent a best case scenario.
The meeting closed with the announcement that the next working group meeting would take place in Tsukuba, Japan during the week of February 8-10, 1999.
|
|
|
|