HIGH THROUGHPUT

1
HIGH THROUGHPUT GENOME ANNOTATION
USING PROTEOMIC TANDEM-MS SPECTRA
Ian Shadforth EngD Research Engineer
Young Bioinformaticians Forum 24th September 2003
2
PROTEOMIC MASS SPECTROMETRY
Digest
MS
MS
3
(No Transcript)
4
(No Transcript)
5
Peptide identified
yes
MS/MS spectra
Mascot
Entry in dB
PHASE 1
no
PHASE 2
Peptide filter
ProbID
General scoring algorithm
Gene- targeted PMF
Entry in dB
Lutefisk
DAS visualisation
Post- processing
PHASE 3
6

• 321 Mascot output files processed
• in 10 hours

• 45 742 peptides matched of which
• 4 264 exceed their 95 confidence score

• 2 581 unique protein ids matched at 95
• confidence level

7

• 2 novel proteins discovered
• CAC3435, score of 287.05
• nucleic acid sequence only
• Q9H111, score of 41.24
• 6 nucleic acid sequence submissions
• and two functional predictions

8

• 41 478 peptide fragments not identified

9

• 41 478 peptide fragments not identified

• Contaminants
• Keratin
• Trypsin
• Polymers

10

• 41 478 peptide fragments not identified

• Contaminants
• Keratin
• Trypsin
• Polymers

• System artefacts
• Poor signal to noise ratio
• Unspecific cleavage
• Unusual fragmentation

11

• 41 478 peptide fragments not identified

• Contaminants
• Keratin
• Trypsin
• Polymers

• System artefacts
• Poor signal to noise ratio
• Unspecific cleavage
• Unusual fragmentation

• PTMs
• Oxidation
• Phosphorylation
• Glycosylation

12

• 41 478 peptide fragments not identified

• Contaminants
• Keratin
• Trypsin
• Polymers

• Novel sequences
• Alternative splice
• variants
• Novel genes

• System artefacts
• Poor signal to noise ratio
• Unspecific cleavage
• Unusual fragmentation

• PTMs
• Oxidation
• Phosphorylation
• Glycosylation

13
(No Transcript)
14

• Retrieval of a gene slice as identified by a
• Mascot probe

• In silico tryptic digest and 3-frame translation

• Mass matching of theoretical to
• experimental peptides

15

• Gene-targeted PMF system appears to
• match any gene to most experiments

• Experiments containing fragments of
• searched genes score more hits

• Lutefisk used to discriminate between
• true and false matches

• Scoring algorithm needs to be developed

16
Peptide identified
yes
MS/MS spectra
Mascot
Entry in dB
PHASE 1
no
PHASE 2
Peptide filter
ProbID
General scoring algorithm
Gene- targeted PMF
Entry in dB
Lutefisk
DAS visualisation
PTM loop
Post- processing
17

• The IDEA-HG DAS server set up by Dr. Steve
  Trevanion

18

• Continued development of the system
• as a BioPipe to process the incoming
• data-stream

• Assessment of the academic communitys
• interest in contributing data to the
• pipeline

• Potential link to a PEDRo compliant
• proteome database

19
QUESTIONS, please