kiekyon's Blog

how current circulating influenza strains could evolve into a new pandemic strain?

search for class specific amino acid mutations conserved in past pandemics, using reverse engineered linear classifiers.

Influenza A has evolved toward host specific mechanisms of infection leading to genetic divergence between human and avian strains

single nucleotide counts are sufficient for classifying the host type

influenza viruses from the pandemics with elements of avian (or avian-like) strains mixed with genetic elements persistent in humans [7-9] are used to provide a historic map of enduring genetic features from past pandemics and their circulation in current human, avian and swine strains

searched for genetic markers conserved in pandemic strains that are associated with two features of infection: host specificity and high mortality rate

New mutations were identified that exhibit a co-variation mutation pattern.

genetic marker identification procedure uses a discriminative classifier

Initial step all single amino acid positions are found that separate the two classes (human/avian or high/low mortality rate)

The iterative step n identifies the n sized (potentially non-contiguous sequence) combinations that separate the data such that each combination does not contain a smaller sized combination that separates the two classes equally well

interactomes of many organisms are far from complete -low interaction coverage -experimental biases

Gene neighbor -Genes with closely related functions encoding potentially interacting proteins are often transcribed as a single unit, an operon, in bacteria and are co-regulated in eukaryotes

Phylogenetic profile -functionally linked and potentially interacting nonhomologous proteins co-evolve and have orthologs in the same subset of fully sequenced organisms

Rosetta Stone -some interacting proteins/domains have homologs in other genomes that are fused into one protein chain

Sequence-based co-evolution -orthologs of coevolving proteins also tend to interact

Classification -use various data sources to train a classifier to distinguish between positive examples of truly interacting protein/domain pairs from the negative examples of non-interacting pairs

Predicting Domain Interactions from Protein Interactions Association -looks for the characteristic sequence or structural motifs which distinguish interacting proteins from non-interacting

Bayesian network models and maximum likelihood -considers other domains in a given pair of interacting proteins -Bayesian network methods treats incorrect and missing interaction data -Maximum Likelihood Estimation method to estimate Bayesian parameter

Domain pair exclusion analysis -accounts for specific, rare interactions between certain members of two domain families -estimating an Eij score which measures the evidence that domains i and j interact and is defined as the logarithm of a ratio of two probabilities

p-Value -tests a null hypothesis that the presence of a particular domain pair in a protein pair has no effect on whether two proteins interact

limitations -domains are assumed to interact independently, although their interactions can depend on other domains in a protein pair -incomplete domain assignments, due to insufficient coverage of domain databases and limited searching ability of domain profiles, can lead to false positive and negative interaction predictions. -protein interaction data is not complete, whereas domain prediction methods are based on this data.