AmiGO: Workflows

From GO Wiki
Revision as of 13:55, 18 December 2006 by Jl242 (talk | contribs)
Jump to navigation Jump to search

AmiGO workflows

Summary diagram

high-throughput experimenter 1

  • name: hte1
  • user: high-throughput experimenter
  • input: list of gene ids, id format in GO database [upload gene list as flat file, or paste into search box)
  • output1: a list of input gene ids with corresponding GO terms [html, download as flat file, XML, Excel?]. Option to filter by evidence code or species

V

  • output2: file above mapped to a set of GO slim terms (GO slim can be chosen from set of existing GO slims or user supplied as a list of ids or a GO format slim file). Option to filter by evidence code or species

V

  • output3: A chart representation of the above slimmed file. Option to filter by evidence code or species

high-throughput experimenter 2

  • name: hte2
  • user: high-throughput experimenter
  • input: list of gene ids, format not in GO database
  • output: same as above [ids invisibly converted by AmiGO using UniProt translation tables?]

microarray 1

  • name: m1
  • user: microarray
  • input: list of 'interesting' gene ids from microarray/proteomics expt, + full gene id set from expt. User specifies id type and species, and which evidence codes to use (default all?)
  • output: List of significantly over-represented GO terms for 'interesting' gene set v/s full set [html, download file]

genetics/genomics user 1

  • name: ev1
  • user: bench scientist, genetics, genomics small-scale or high-throughput [upload as file or paste into search box. List of options - find matches via BLAST, InterProScan, others?]
  • input: list of sequences (protein or DNA)
  • output1a: BLAST results [html + file download option]. How should this be displayed? Default to top GO term, with Option for show alignments, show top 10 hits? Should we allow users to configure the hits? Should ultimately end up with a list of sequence ids with GO terms

OR

  • output1b: InterProScan output [we'd have to look into what their output is but, ultimately we want them to be able to download a seq id->GO term file]

V

  • output2: slimmed version of above file [as before]

V

  • output3: A chart representation of the above slimmed file [as before]

bench research biologist 1

  • name: brb1
  • user: bench research biologist
  • input: single gene id, format in GO database
  • output: all annotations for that term [as it is now]

bench research biologist 2

  • name: brb2
  • user: bench research biologist
  • input: single gene id, format not in GO database
  • output: as above, but we need to convert id behind the scenes

bench research biologist 3

  • name q1
  • user: bench research biologist
  • input: specific biological query e.g. show me all the genes involved in the cell cycle that are not kinases, or give me all mouse genes that have anything to do with chromatin
  • output: A list of the genes/proteins matching the query, perhaps also broken down by GO term. Option to filter by evidence code or species.