David

Looking at the possibility of using David instead of or in addition to IDecoder.

We can download text files from here: Request Knowledgebase Files. You have to log in to download the files.

You can choose the species and a central identifier. Also you can check some annotation categories. I have been checking all of General Annotations and Main_Accessions. You are allowed to check a maximum of 10 annotation categories. So I checked the following Annotations:

1. Chromosome
2. Cytoband
3. Entrez_gene_summary
4. Homologous_gene
5. Official_Gene_Symbol
6. PIR_Summary
7. SP_Comment
8. ENSEMBL_GENE_ID
9. ENTREZ_GENE_ID

After clicking Submit you get an email that allows you to download a zip file containing more files.

For example, when I chose the Central Identifier as "AFFYMETRIX_EXON_GENE_ID", I received the following files:

1. AFFYMETRIX_EXON_GENE_ID2CHROMOSOME.txt
2. AFFYMETRIX_EXON_GENE_ID2CYTOBAND.txt
3. AFFYMETRIX_EXON_GENE_ID2ENTREZ_GENE_SUMMARY.txt
4. AFFYMETRIX_EXON_GENE_ID2HOMOLOGOUS_GENE.txt
5. AFFYMETRIX_EXON_GENE_ID2OFFICIAL_GENE_SYMBOL.txt
6. AFFYMETRIX_EXON_GENE_ID2PIR_SUMMARY.txt
7. AFFYMETRIX_EXON_GENE_ID2SP_COMMENT.txt
8. AFFYMETRIX_EXON_GENE_ID2ENSEMBL_GENE_ID.txt
9. AFFYMETRIX_EXON_GENE_ID2ENTREZ_GENE_ID.txt
10. AFFYMETRIX_EXON_GENE_ID2DAVID_GENE_NAME.txt
11. AFFYMETRIX_EXON_GENE_ID2TAX_ID.txt

Note that there are two extra files - 11 files and 9 requested annotations.

The DAVID_GENE_NAME file is interesting. It has, as the second column, a name for the gene. However, the same name can match multiple AFFYMETRIX ID's. I view this as identifying the "connected component". That is, it specifies a unique gene that can have many other ID's.

How could we use these files?

• Download files for every possible central identifier.
  • Perhaps not every possible central identifier.
• Create a table DAVID_IDENTIFIERS with two columns, each unique

1. DAVID_GENE_NAME varchar2
2. DAVID_ID Number

• Create a table OTHER_IDENTIFIERS with four columns.

1. identifier varchar2
2. identifier_Source varchar2
3. id_number number
4. david_id number The two columns identifier and identifier_source will together be unique. id_number will be unique. There may be many records with the same david_id. Those records with the same david_id should be considered to be the same gene.

• We'll have to figure out what to do if a user enters an ID for a gene that is not unique and that corresponds to multiple david_ids.
• But the basic idea will be that genes will have a unique david_id.
• From the david id, we can get all other corresponding id's without a search.

List of all possible ID sources:

1. AFFYMETRIX_3PRIME_IVT_ID
2. AFFYMETRIX_EXON_GENE_ID
3. AFFYMETRIX_SNP_ID
4. AGILENT_CHIP_ID
5. AGILENT_ID
6. AGILENT_OLIGO_ID
7. ENSEMBL_GENE_ID
8. ENSEMBL_TRANSCRIPT_ID
9. ENTREZ_GENE_ID
10. FLYBASE_GENE_ID
11. FLYBASE_TRANSCRIPT_ID
12. GENBANK_ACCESSION
13. GENOMIC_GI_ACCESSION
14. GENPEPT_ACCESSION
15. ILLUMINA_ID
16. IPI_ID
17. MGI_ID
18. OFFICIAL_GENE_SYMBOL
19. PFAM_ID
20. PIR_ID
21. PROTEIN_GI_ACCESSION
22. REFSEQ_GENOMIC
23. REFSEQ_MRNA
24. REFSEQ_PROTEIN
25. REFSEQ_RNA
26. RGD_ID
27. SGD_ID
28. TAIR_ID
29. UCSC_GENE_ID
30. UNIGENE
31. UNIPROT_ACCESSION
32. UNIPROT_ID
33. UNIREF100_ID
34. WORMBASE_GENE_ID
35. WORMPEP_ID
36. ZFIN_ID

Go Back to Future Directions