The fuzzy inference is achieved by the Min and Max operations.
Inference is achieved by formulating an appropriate MCMC algorithm that samples from (Gamerman and Lopes, 2006).
The joint inference is achieved by using a multi-population group lasso (MPGL), with an L1/ L2 regression penalty (Obozinski et al., 2008; Yuan and Lin, 2006; Zhao et al., 2008) that encourages (but does not enforce) multiple populations to have similar causal markers.
Inter-species sequence conservation inference is achieved by performing an all-against-all search over all the peptide sequences of all the species of an order (as given by organismal taxonomy of RefSeq), considering a sequence identity and coverage of >50% each.
A good compromise between low percentage of false positives (high inference success) and a good statistics was achieved at T0 = 1.50 for E. coli, and T0 = 1.30 for B. subtilis.
The identification of codon sites under positive selection was achieved through Bayesian inference using the Selecton2.2 server [ 75] with the "Positive selection enabled (M8, beta + w > =1)" evolutionary model with 8 categories, on alignments of S. sclerotiorum 1980, B. cinerea b05.10 and B. cinerea t4 orthologs.
Then, the probability inference could be achieved using inference algorithm, such as junction tree method [ 25, 26].
Validation of the tree topology was achieved through multiple methods of phylogenetic inference.
We consider this hypothesis is reasonable since it is widely assumed that similar phenotypes may be caused by functionally close related genes [ 15, 16], thus if more information about protein-protein interactions, gene-phenotype associations as well as phenotype-phenotype similarities is known, higher inference accuracy in gene-phenotype relationship inference will be achieved.
With randomly selected six-digit numeric strings of four different scripts; an average of 96.72% script inference accuracy is achieved.
