Semantic resources project/Use Cases/Meeting Notes/Jang-Ho Cha/Meeting 09292009

JHC Meeting Notes 9/29/2009

In attendance: JHC, KMcF, JAR, TWD

1. JHC demos CHDI X-roads software system
 * scores: categorical, ranked values.
 * references

2. JHC: talking about microarray data, ChIP-*: the problem is that a lot of data is “lost,” an experiment is performed, but most of the results disappear into papers or results sections and aren’t used again. They’ve mostly dealt with gene expression but are looking to move into other experiment types: ChIP-seq, etc. How do they combine these with their own data?
 * he’s presenting their own computational problems.
 * most of this is in the context of analyzing the high-throughput genome-scale datasets.
 * JAR and I are talking about data integration problems *after* this analysis (in some sense)

3. JAR and TWD: talk about the purpose of the SCF/SC collaboration project, describe SWAN
 * demo SWAN for JHC
 * JAR talks about ontologies as a “next step,” a way to combine your results with other datasets.
 * TWD thinks about it as a way of combining datasets for a joint analysis (post- vs. pre-processing).

4. TWD: asks KM about her typical workflow –
 * Proteins -> antibodies -> “interactors” -> publications (or -> expression data through publications)
 * She has used some pre-packaged analysis systems on the web:
 * GSEA (From the Broad Institute) : http://www.broadinstitute.org/gsea/
 * Genomatix Bibliosphere: http://www.sigmaplot.com/products/genomatix/bibliosphere.php
 * NCBI Interactors: HPRD http://www.hprd.org/
 * Often, starting from a set of expression experiments, the next questions are ...
 * What mouse strain was used?
 * What transgene was used?
 * What tissue? etc.

5. KM describes some websites that she uses:
 * biocompare.com: http://www.biocompare.com/
 * Links antibodies to suppliers
 * Contains forum spaces, where users can give advice about experiments, antibodies, protocols
 * Reviews from users
 * Entrez Gene/ Pubmed
 * EMBL
 * Scientist Solutions: http://www.scientistsolutions.com/
 * She uses the protocols forum...
 * Biotechniques: http://www.biotechniques.com/
 * She doesn't read methods journals or articles as much as she uses these sites.

6. JHC describes more "nagging suspicions" as they think about analyzing high-throughput datasets
 * More information ("dark" information) can be pulled out through informatics ("Are we throwing data away?")
 * Biomarker development (biomarker resources) are an important step.