Minutes for Working Group Meeting, February 4, 2004, Tsukuba, Japan

Executive Summary:

The IRGSP is easily on pace to complete the finished quality sequencing of the rice genome by December, 2004. The Table below indicates the number of clones in the minimum tiling path, numbers submitted to the PLN section of GenBank as of February 1, 2004, the rate of finishing over the last nine months, and the remaining physical gaps. These totals are shown for the groups actively engaged in finished sequencing and do not reflect the contributions of all of the members of the IRGSP. It is estimated that the current physical map covers 90% of the genome and of that 71% of the coverage is finished based on non-overlapping nucleotides.

The IRGSP is conducting two initiatives to verify sequence quality. A mechanism is in place to check sample finished clones submitted to public databases. An automated utility has been implemented at AGI that compares in silico restriction enzyme digests if submitted clones with HindIII digests used to prepare the the physical map.

The IRGSP is close to submitting a joint paper on the current sequence and in this work has been aided by external experts with aspects of genome analysis.

Progress Reports:

Taiwan: Han-Hwa CHEN reported that they had they had 87 clones to finish on chromosome 5 not counting the 79 that they are sharing with AGI/CSHL. They are currently finishing at the rate of about 9 clones per month which means that at the current rate they should finish in time. Dr. CHEN mentioned that they had between 10 and 20 difficult BACs and were considering sending them to AGI/CSHL as well. Rod WING urged them to do so as quickly as possible.

There are five gaps on chromosome 5. Dr. CHEN said that they had located three clones that covered the centromere and that they were continuing to analyze the remaining gaps with fiber FISH.

ACWW: Rod WING reported on the 8 gaps that remain in regions claimed by ACWW. Of these, one on chromosome 3 has been closed. Of the remaining two on chromosome 3, one has been measured to be 50 kb and the other at greater than 500 kb.

ACWW is working in collaboration with Taiwan, India, Korea and PGIR to finish some or all of their unfinished clones. Rod also reported that for the 79 clones from chromosome 5, the 46 from chromosome 11, the 15 clones from chromosome 9, and the remainder of the unfinished clones from PGIR, that they are sequencing in collaboration, that things are well in hand.

China: Bin HAN reported that of the 19 unfinished clones on chromosome 4, 16 were down to 1 or 2 sequencing gaps and that the other three had less than 10 gaps. There are 4 physical gaps on chromosome 4. One of these is more than 160 kb. They are doing fiber FISH to confirm the size of the remainder.

France: Francis QUETIER reported that of the 279 clones on chromosome 12 (including 11PCR products and one BAC sequenced by PGIR) 235 BACs and 9 PCR products were finished and 32 clones and 2 PCR products were under finishing. 221 clones had been validated and that 5 were under validation. Of the remaining clones that are in finishing 16 have problems with the number an orientation of repeats and 23 have microsatellite problems.

TIGR: Robin BUELL reported 121 clones in closure. Of the 7 remaining gaps in their claimed regions, two include centromeres on 3 and 11. Two non-centromere gaps on chromosome 3 are estimated at less than 100 kb and these will be measured by fiber FISH.

India: Saurabh RAGHUVANSHI reported that had identified 4 additional gap filling clones to add to their minimum tiling path. They have 25 unfinished clones to finish themselves and at the current rate of 3.5 per month they anticipate no difficulties.

Japan: Takashi MATSUMOTO spoke for the RGP and presented graphical and tabular representation for progress of the RGP and the IRGSP as a whole. The RGP has completed 72.6% of the nucleotides to finished level for the estimated 204.7 MB that they have claimed. Overall, the IRGSP has completed 71.4% of an estimated 398.3 MB to the finished level.

Takashi also gave a presentation of the use of fiber FISH technology with examples of the measurement of a 415 kb gap on the long arm of chromosome 8 and the measurement of a 30 kb subtelomeric gap also on 8L.

Finishing Methods and Gap Filling:

Dick MCCOMBIE and his colleagues from Cold Spring Harbor Laboratories conducted a Finishing Workshop on February 2 and 3 at the RGP in Tsukuba. It was attended by eight students. The workshop covered the basics of finishing, finishing project organization, and new ways to solve difficult finishing problems. Students were urged to bring difficult problems and reports were that new ways of attacking these difficult clones were learned. The course was aimed at both advanced and beginning finishers. Basic tools were taught, but advanced finishers were also be able to share their experience and knowledge. Attendees reported the course to have been very profitable.

Jianzhong WU presented approaches to gap filling used at the RGP and discussed a proposal to use YAC inserts for the approximately 1/5 of the gaps that are spanned by these clones. Dick MCCOMBIE asked if he had considered using YACs a sequencing template and offered to write to the Sanger Center for their protocols.

Jianzhong described his collaboration with AGI to genetically map a set of the largest of the unanchored BAC contigs. He said that to start they had performed BAC end sequencing to provide unique sequences for mapping.

Rod WING agreed that it would be a good idea to get better BAC end sequences from the unanchored BACs. He mentioned that his colleague Yeisoo YU is draft sequencing unanchored contigs with the view to finding sequences that fill gaps. He urged Jianzhong to visit Tucson to see how they could work together on this problem.

Dick MCCOMBIE mentioned that his group had learned that they were able able to successfully sequence an entire contig of 18 BACs and that sequence to 0.75 X coverage would give them enough sequencing information to learn whether the contig was worth pursuing.

Quality Control Proposal:

Dick MCOMBIE has refined his proposal to randomly check the quality of submitted finished IRGSP BACs/PACs. During the subsequent discussion the following protocol was accepted by all members. He prefaced his remarks by saying that recent experience had taught him that with modern sequencing technology, problems in rice would most likely be in assemblies rather than in sequencing mistakes.

1) Ten finished clones from each of the seven groups that is doing finished sequencing will be chosen at random from GenBank. Each group will be asked to provide BAC/PAC DNA and stabs for each of the clones.

2) Multiple digests will be performed at AGI on the BAC DNA. Enzymes will be chosen based on the size of fragments that are predicted from the sequence. The sizes of the fragments in the digests will be automatically scored and compared with in silico digests.

3) Based on these results, three of the clones will be chosen for further study where assemblies will be checked at both CSHL and the RGP. Each group will requested to download the assemblies and trace files for these clones to an FTP site and to send both a BAC/PAC stab and the clone DNA.

4) A report on the results will be sent to the originating group and a response to the report will be expected.

It is anticipated that it will take a month to request and receive material and an additional month the check the assemblies.

It was agreed that this was the first step in the process and that the Working Group would evaluate the results before proceeding to the next step. If there were problems with assemblies from any specific group, more clones will be checked by actual and in silico digests.

In fact, all CUGI BAC clones sequenced and submitted from the IRGSP are currently subjected to a first pass check at AGI. As clones are submitted their sequences are automatically downloaded and subjected to an in silicoHinDIII digest which is compared with the FPC digests on record. On the basis of this result the clone is checked for assembly into existing contigs to learn whether it reassembles into the same previously assigned contig and for the goodness of the level of fit. A preliminary look indicates that all finished clones appear to reassemble in their original contigs and most do with very low error scores. This calculation breaks down when only the non-overlapping portion of a clone is submitted. The results for many clones are available at http://www.genome.arizona.edu/cgi-bin/status/rice/status.cgi (look at FPC Contig and Confidence Score columns for each chromosome).

webmaster@staff.or.jp Copyright (C) The International Rice Genome Sequencing Project (IRGSP). 2005 All rights reserved.