1. Microarrays: Goal — to make technique accessible to as broad a group as possible: one reader per 20 active-user laboratories; for laboratories that are making their own arrays, one per 20 active-user laboratories.

We endorse the academic use as well as development of the technique. This can best be accomplished at present by the method of polynucleotide arrays (Pat Brown/Synteni).

There was considerable interest by some people for purchasing ready-made arrays. The manner in which the NIH would subsidize this is unclear. The cost per array should ideally be no more than $50-100 per array.

We recommend that the NIH support, perhaps via instrumentation grants, the purchase of arrayer (robot for spotting) and reader. We appreciate that these pieces of equipment are under development but feel that the yeast community has immediate use of this technique and that their use will lead to improvements in the technique.

Approximate cost for arrayer: $50,000

Approximate cost for reader: $50,000-100,000

Cost per array: at present, primer pairs for amplifying all yeast ORFs cost approximately $30,000 and will produce ca. 1000 arrays, for a cost per array of $30.

Arrays need a set of standard controls to allow comparison from experiment to experiment and from laboratory to laboratory.

2. Mass spectrometric identification of components of protein complexes.

We anticipate that the demand for mass spectrometric analysis will increase as genome sequence information increases. We therefore recommend the establishment of a small number of national centers to which samples from any number of organisms are sent for analysis.

Approximate cost for instrumentation: $1,500,000

Personnel cost (operator/technician): $75,000 annually

3. Deconvolution microscope: approximately 1 per 3-4 user laboratories.

Deconvolution microscopy is particularly critical for analysis of small cells such as yeast and Chlamydomonas. It is preferable to confocal microscopy because it can take thin optical sections and tilt the sample to assess co-localization. In addition, the software for image analysis is able to remove out-of-focus information and produce sharper images.

Approximate cost: $150,000 for 3-I; $300,000 for DeltaVision.

Purchase of such microscopes is not expected to be funded through RO1 grants.

 

1. Heterologous cDNA libraries for regulated expression in yeast: care needs to be given to design of the vector and to produce full-length cDNA. Libraries should be made from human and mouse, as well as Xenopus, and other model organisms.

Approximate cost: $75,000-$100,000 for one-two years.

Construction of such libraries might be coordinated with ongoing efforts to construct full-length cDNA libraries for CGAP and mouse GAP projects.

2. Unigene set of plasmids carrying each ORF. This set can be used for: (a) cloning by complementation, (b) facile construction of derivative sets allowing inducible expression, overexpression, epitope-tagged, etc. in a variety of vectors (integrating, high copy, etc.), and (c) amplification of ORFs for array construction.

The initial unigene set of plasmids can in principle be converted to the variant sets using, for the example, the method of Elledge and colleagues.

Approximate cost: $50,000 for construction of unigene set.

3. Chemical libraries. Such libraries could be used for: (a) determining the response pattern of cells to drugs, (b) determining sensitivity of given strains to drugs, and (c) carrying out medium-scale (nonrobotic) screening of mutants with the libraries.

These libraries would be used in the laboratories of the investigators and supplied in 96-well plates.

General comment about distribution of genomic resources:

A mechanism for distributing genomic resources such as the unigene set of plasmids and heterologous cDNA libraries at nominal cost needs to be developed. The methods for distributing such material to the yeast community will be a preview of what will be required for distribution of comparable materials to researchers working on other model organisms. Distribution of genomic resources through commercial channels is currently prohibitively expensive for allowing full access to many important reagents, for example, those necessary to produce DNA microarrays.

DATABASES

1. Importance: The Saccharomyces Genome Database was recognized as being enormously successful and essential not only for the yeast community but valuable for others as well. The successful functioning of SGD leads to additional suggestions for ways in which its usefulness can be expanded. SGD is in many respects a model for genome databases.

2. Proposals:

a) The SGD should be a repository for unpublished information on mutant phenotypes.

b) The SGD should have more effective links to other organism databases.

c) The SGD should consider curating fission yeast.

d) The SGD should curate pathways such as metabolic, signal transduction, etc.

e) SGD should consider curating and cross-referencing expression array data and developing methods for sorting existing array data. We recognize this as a major challenge, but it is an extraordinarily important undertaking.

f) The SGD should consider developing a phenotype-based search engine.

g) The SGD should provide image data for protein localization or at least link to the Yale Genome Center.

h) Attempts should be made to give different organism databases the same look and feel.

1. Chlamydomonas: unique feature — exceptional opportunity to study structure of flagella (ca. 250 proteins): wish list — physical map and EST.

2. Schizosaccharomyces pombe: has many functional and structural differences from S. cerevisiae but has many of the same advantageous technical features (which can be enhanced). Top priority request: capability of microarrays; next: enhanced database with annotation. Fission yeast should be included in various yeast initiatives.

3. Neurospora: rationale: the filamentous fungi have a dramatically different life style from budding yeast; this is reflected currently in its EST sequences, many of which are novel.

A variety of filamentous fungi were discussed in addition to Neurospora, including Aspergillus nidulans (a pathogen of immunocompromised individuals) and Ashbya gossypii (a cotton pathogen whose genome appears related to budding yeast prior to duplication).

One criterion for choosing a filamentous fungus for conferring expanded genomic resources is that it have homologous recombination to facilitate genetic analysis.

4. Model organisms for infectious disease: several infectious agents exist that have great potential as model organisms because of their relevance to important infectious diseases but are experimentally manipulable. These organisms include Toxoplasma, Candida albicans, Cryptococcus neoformans, Ustilago maydis (corn smut), and Histoplasma capsulatum. As noted above for Neurospora and other filamentous fungi, an important criterion for choosing model infectious disease organisms for expanded genomic resources is that it have homologous recombination to facilitate genetic analysis.