Interpreting consensus sequences based on plurality rule (Q1205295): Difference between revisions

The authors study a special case of their paper ``Consensus sequences based on plurality rules'' reviewed above. If \(P\) is a profile of alternatives (from \(S=\{A,C,G,T\})\), denote by \(\Gamma(P)\) the set of alternatives which occur in \(P\); e.g. if \(P=(A,A,A,C,A,G,A,C)\) then \(\Gamma(p)=\{A,C,G\}\). A profile \(P\) is called balanced if the \(|\Gamma(P)|\) different alternatives in it all occur with the same frequency (or, if this is impossible, differ by at most one occurrence); e.g. \(P=(A,A,A,C,C,C,T,T,T)\) and \(P=(A,A,A,C,C,G,G)\) are balanced. A balanced profile \(Q\) is a closest balanced profile for a given profile \(P\) (of the same length \(k)\) if the number of cases where \(p_ i\neq q_ i\) is minimal (among all balanced profiles \(Q)\). The plurality rule \({\mathbf p}\) associates with a profile \(P\) the collection of subsets \(\Gamma(Q)\), where \(Q\) ranges over the closest balanced profiles for \(P\); e.g. if \(P=(A,A,A,A,G,G)\) then the closest balanced profile is unique and given by \(Q=(A,A,A,G,G,G)\), so that \({\mathbf p}(P)=\{A,G\}\). The authors then analyze this consensus function numerically. By expressing the plurality rule as a composition of simpler functions, they obtain both an algorithm to calculate the consensus result and an upper bound on the number of nonequivalent results. For example, when used to analyze molecular sequences such as DNA or RNA, the plurality rule yields at most 48 nonequivalent results. This number is small enough to enable biological researchers to interpret results obtained by this method.