Email updates

Keep up to date with the latest news and content from Genome Biology and BioMed Central.

Open Access Highly Accessed Research

Protein structure protection commits gene expression patterns

Jianping Chen1, Han Liang2 and Ariel Fernández134*

Author Affiliations

1 Program in Applied Physics, Rice Quantum Institute, Rice University, Houston, TX 77005, USA

2 Department of Ecology and Evolution, University of Chicago, Chicago, IL 60637, USA

3 Department of Bioengineering, Rice University, Houston, TX 77005, USA

4 Department of Computer Science, University of Chicago, Chicago, IL 60637, USA

For all author emails, please log on.

Genome Biology 2008, 9:R107  doi:10.1186/gb-2008-9-7-r107

Published: 7 July 2008

Additional files

Additional data file 1:

Data in column A indicate the expression correlation η associated with protein interactions, and data in column B indicate the structure vulnerability ν for interactions within specific complexes. The rest of the columns contain the ORF, domain and structure information (PDB accession code of interacting domain or its Pfam-homologs), respectively, for every pair of interacting proteins.

Format: XLS Size: 38KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional data file 2:

Sheet 1 contains all Pfam-filtered yeast protein interactions, while sheet 2 contains only those interactions with both partners having PDB structures. In each sheet, column A lists the expression correlation h of interactions, and columns B and C list the structure vulnerability n of interactions not involving or involving, respectively, extremely vulnerable proteins. The remaining columns contain ORF, domain and structure information (PDB accession code of interacting domain or of its Pfam-homologs) for every pair of interacting proteins.

Format: XLS Size: 351KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional data file 3:

Sheet 1 contains all Pfam-filtered human protein interactions, while sheet 2 contains only those interactions with both partners having PDB structures. In each sheet, column A contains the expression correlation h for each interaction, and columns B and C list the structure vulnerability n of interactions not involving or involving, respectively, extremely vulnerable proteins, and the rest of the columns list gene name, protein ID, domain and structure information (PDB accession code of interacting domain or of its Pfam-homologs) of every pair of interacting proteins.

Format: XLS Size: 128KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional data file 4:

The extremely vulnerable proteins in human are identified from genome-wide scanning of protein-encoding regions with sequence windows (length ≥ 30) containing mainly amino acids (G, A, S, Y, N, Q, P) that are poor protectors of the protein backbone. An extremely vulnerable protein contains at least one such window with a threshold of three amino acids allowed to be outside the group of poor protectors.

Format: XLS Size: 34KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional data file 5:

Extremely vulnerable yeast proteins are determined in the same way as for human (Additional data file 4). The rows marked in green correspond to the five confirmed yeast prions [26-29]: SUP35 (ERF2), URE2, NEW1, RNQ1 and SWI1.

Format: XLS Size: 26KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional data file 6:

The number of putative target-site types corresponding to 162 conserved miRNA families determined for 17,444 human genes by interrogation of the 3' UTR using TargetScanS (version 4.0) [45].

Format: XLS Size: 1MB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional data file 7:

The co-expression similarity for genes i, j, encoding a pair of interacting proteins is alternatively measured as the adjacency aij(β) = (0.5 + 0.5 η (i, j))β, where η (i, j) is the expression correlation for the gene pair i, j and β is a soft threshold [46]. Similarly, the structure vulnerability is alternatively defined as νi, j(β) = ν (i, j)β, where ν (i, j) is the maximum ν-value for the interacting pair. (a, b) (β)-a(β)) correlations for yeast for exponents β = 0.5 (a) and 10 (b). The adjacencies for β = 1 correspond simply to a linear rescaling of η already correlated with ν in Figure 4. (c, d) The same as (a, b) but for human. Notice that high exponents (β > 1) tend to amplify differences in co-expression, yielding lower correlation coefficients (R2 in (ν (β)-a(β)) plots).

Format: EPS Size: 4.8MB Download file

Open Data

Additional data file 8:

A normalized co-expression similarity γ (β, complex) for all genes encoding proteins that form a complex is obtained from the adjacencies of the pairwise interactions within the complex as: γ (β, complex) = [mediani, j ∈ complexaij(β)]/mediani, j aij(β)], where the median in the denominator extends over all interactive pairs in the interactome. Similarly, the normalized structure vulnerability Λ (β, complex) for complexes is defined as Λ (β, complex) = [mediani, j ∈ complexνij(β)]/mediani, j νij(β)]. (a-c) (β, complex)-γ (β, complex)) correlation over all 98 yeast complexes with transcriptome representation for exponents β = 0.5 (a), 1 (b) and 10 (c). (d-f) The same as (a-c) but for 53 human complexes.

Format: EPS Size: 5.2MB Download file

Open Data

Additional data file 9:

(a) Yeast complexes. (b) Human complexes.

Format: XLS Size: 51KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional data file 10:

(η-ν) plot obtained for the yeast developmental-phase transcriptome obtained from a comprehensive identification of cell cycle-regulated genes by microarray hybridization [23]

Format: EPS Size: 1.5MB Download file

Open Data