CrossSpecies Proteomic Identification

1
Cross-Species Proteomic Identification

• Sheffield University
• 17th May 2004

2
GAPSIA Proteomics

• Cross-Species Proteomics
• Aspergillus Protein Website

Chris Cole - BmS UMIST - 2004
3
Cross-Species Identification

• Using PepMapper
• Identifies proteins from the masses of peptide
  fragments (Peptide Mass Fingerprinting, PMF).
• Cross-species identification is possible with
  sequence identities of 70 using std PMF.
• Do protein profiles instead of single orthologues
  make a difference?

Chris Cole - BmS UMIST - 2004
4
Hits of A. nidulans Orthologues using Std PMF
(50ppm, 500Da cut-off)
291 orthologues against A. fumigatus database
Chris Cole - BmS UMIST - 2004
5
Effect of MS Accuracy - A. nidulans (14peps,
500Da cut-off)
Chris Cole - BmS UMIST - 2004
6
Multiple Genomes - A. niger(14peps, 10ppm, 500Da
cut-off)
Chris Cole - BmS UMIST - 2004
7
A. niger Orthologues in the Fungi
Chris Cole - BmS UMIST - 2004
8
PMF Profile Analysis
Pep1
Pep2
Pep3
A. nidulans
A. fumigatus
A. niger
A. nother
An. other
Chris Cole - BmS UMIST - 2004
9
Limitations of Standard Approach

• One Protein - One Orthologue (per species)
• Ignores other related sequences
• Potential for false positives
• Not every protein has a putative orthologue
• but, may have several homologues

Chris Cole - BmS UMIST - 2004
10
Profile Analysis

• One Protein - One Profile
• PSI-BLAST finds gt1 homologue per search protein
• 3592 profile hits vs 876 orthologue hits (6 Fungi
  2 Asp)
• Optimisation of search parameters critical
• Sequence similarity cut-off is useful
• excluded all profile sequences with lt60 id.

Chris Cole - BmS UMIST - 2004
11

• Standard PepMapper
• HITgt 001 asfu07167 1.00e-00 56.21 33.75
  0.00 6 from 14 42.9 pr00022 oasni00128
• HITgt 002 asnu00413 7.95e-09 37.56 22.49
  0.00 4 from 14 28.6 pr00022 oasni00128
• HITgt 003 fugr09715 6.97e-09 37.43 22.62
  0.00 4 from 14 28.6 pr00022 oasni00128
• HITgt 010 necr03666 2.96e-17 18.15 11.07
  0.00 2 from 14 14.3 pr00022 oasni00128
• Profile Mode
• PROgt 00001 00022 tot 158.9229 av 31.7846
  seqs 5 av2 31.7846 seqs 5
• PROgt 00002 00045 tot 27.0262 av 3.8609
  seqs 7 av2 9.0087 seqs 3
• PROgt 00003 00179 tot 16.7540 av 2.3934
  seqs 7 av2 8.3770 seqs 2
• PROgt 00004 00065 tot 18.1873 av 2.0208
  seqs 9 av2 9.0936 seqs 2
• PROgt 00005 00158 tot 8.0236 av 2.0059
  seqs 4 av2 8.0236 seqs 1
• PROgt 00006 00077 tot 9.9925 av 1.9985
  seqs 5 av2 9.9925 seqs 1
• PROgt 00007 00026 tot 7.9422 av 1.9855
  seqs 4 av2 7.9422 seqs 1
• PROgt 00008 00107 tot 9.7987 av 1.9597
  seqs 5 av2 9.7987 seqs 1
• PROgt 00008 00002 tot 9.7987 av 1.9597
  seqs 5 av2 9.7987 seqs 1
• PROgt 00010 00017 tot 55.7179 av 1.7974
  seqs 31 av2 9.2863 seqs 6

Chris Cole - BmS UMIST - 2004
12
Profile Searching of A. niger Orthologues
(14peps, 500Da cut-off, 10ppm)
13
Parameter Optimisation

• Query set of 198 UniProt A. niger sequences.
• Against a filamentous fungi database
• A. nidulans
• A. fumigatus
• N. crassa
• M. grisea
• F. graminearum

14
Hits with Varying PSI-BLAST e-values(4
iterations)
15
Trends in Profile Make-up with Varying PSI-BLAST
e-values
16
lt60 id Profile Sequences Removed(4 iterations)
17
Trends In Profile Make-up at 60 Seq. id Cut-off
18
Size of Profiles
19
Aspergillus Database Website Update

• http//wolf.bms.umist.ac.uk/mjfsscc/GAPSIA/entry.
  html

Chris Cole - BmS UMIST - 2004
20
Whats in the DB?

• A. nidulans
• 9855 sequences (sptr Broad)
• 184,836 orthologues
• 114,373 paralogues
• A. fumigatus
• 11,145 sequences (sptr Sanger)
• 150,832 orthologues
• 47,898 paralogues

• A. niger
• 14,530 sequences (sptr GCI)
• 193,338 orthologues
• 126,113 paralogues
• Annotations
• 1074 from sptr
• 6530 from GCI
• 151 from Broad

Chris Cole - BmS UMIST - 2004
21
Website - Whats New?

• Tidied up search page
• Tidied up protein detail page
• Improved linking in BLAST output
• New PepMapper interface

Chris Cole - BmS UMIST - 2004
22
Chris Cole - BmS UMIST - 2004
23
Chris Cole - BmS UMIST - 2004
24
Chris Cole - BmS UMIST - 2004
25
Where to next?

• Tandem MS data searching
• Against existing dbs (done kinda)
• CADRE (can do, easy)
• raw genomes (could do, difficult)
• Website
• Crossrefs to CADRE
• More cross-species views
• Cross-species proteomics tools
• PMFbased
• Tandem

Chris Cole - BmS UMIST - 2004
26
Acknowledgements

• Simon Hubbard
• Steve Oliver
• Patrick Lester
• BBSRC

http//wolf.bms.umist.ac.uk/mjfsscc/GAPSIA/entry.
html
Chris Cole - BmS UMIST - 2004