Shotgun proteonomics
CREATED: 200808140219 Speaker: Christian von Mering
** Proteonomics workflow
peptide sequencing
identify proteins
spectral counting
abundance levels
- About 50% of proteins are detected in current workflow
- Detailed qualitative characterization of proteins not identified in C. elegans
- Improved genome annotation (new start site, alternative splicing)
** Spectral counting: quantitative information of protein abundance
- convectional methods relay on peptide labeling, measure relative changes
- spectral counting: directly from shotgun data, no need for peptide labels
- problem: not all peptides are produced at cleavage points, not all peptides are detected by the machine
- very short/long peptides are not observed
- $a = (\sum_i count(p_i) \times length(p_i))/(\sum_j length(q_j) \times f(q_j))$
- p = identified peptides, q = tryptic peptides, f(q) = peptide length correction factor
- comparison of worm with yeast
- genes which are part of operons in C. elegans are more abundant and more easily detected
- operons are probably not cause for abundance, because orthologous genes in fly are also more abundant
** Comparative analysis
- orthologous protein pairs in worm and fly
- correlation of 0.79 between abundance of orthologs in worm and fly (2691 pairs of orthologs)
- breakdown of correlation of abundance levels according to functional categories
- protein abundance have better correlation compared to comparison of transcript abundance => mRNA can drift
- protein vs transcript within an organism
** Future work
- multiple species
- protein networks
- comparative approach to PTM