Molecular biology is organized hierarchically: proteins form complexes which create pathways and meta-pathways that influence phenotype. Characterizing this hierarchy in a general way is difficult due to the complexity of genomes.

SCALES is a framework that analyzes patterns of co-variation amongst genomic features across a diversity of bacteria to infer a hierarchy of statistical interactions between proteins. The statistically derived hierarchy of molecular interactions reflects the hierarchy of biological function across different pathways in different bacterial organisms. In brief, given a genotype, SCALES provides a hierarchy of protein interactions that are relevant to understanding phenotype.



Services Provided:

Interactions: Pre-Computed

Precomputed indirect (shared functional) and direct (physical) protein-protein interactions predicted using the approach detailed in the manuscript are provided for 7047 bacterial reference proteomes present in the UniProt database release 02/2020. All organisms are searchable through a drop-down menu and results can be downloaded as .csv or .pdf formats. If specific interactions are desired, protein query can also be included using either protein name or gene name (i.e. ‘Motility protein B’ or ‘motB’).

Pre-Computed Proteomes

Interactions: De Novo

Given a bacterial sequence (.fasta file) of interest uploaded by a user, proteome-wide predicted indirect and direct interactions are provided. Results can be downloaded as .csv or .pdf formats.

De Novo Proteomes

Pathway Organization

SCALES provides a layered, hierarchical set of protein interactions that describe networks of protein interactions encoding detailed functions, broad functions, and pathway-level functions given a query protein of interest. This type of hierarchical architecture is derived using a query protein of interest (PFAM ID) and is visually represented as a network graph that can be dynamically modulated according to a metric termed ‘spectral depth’. Shallow spectral depths reflect global biological functions while deeper depths reflect the specific functions that create the global functions. Here, we provide the ability to create such hierarchical diagrams for a query protein of interest in a bacteria. Contours placed around modules of statistically interacting proteins are related to ontological terms provided by Gene Set Enrichment Analysis (https://david.ncifcrf.gov/home.jsp) for aiding in interpretation of the graph. Graphs can be created for any query protein within either a bacterial reference proteome found in UniProt or a sequenced bacterial genome uploaded in .fasta format. Pairwise spectral depth matrices used to generate the hierarchical graphs can be downloaded as .csv or .pdf formats. Network graphs can be downloaded as vector graphic (.svg).

Pathway Organization