T-Coffee

T-Coffee (Tree-based Consistency Objective Function For alignment Evaluation) is a consistency-based multiple sequence alignment method developed by Cedric Notredame in 2000. Unlike progressive aligners like Clustal, T-Coffee first aligns all pairs of sequences, then uses the transitivity of alignment to build a consistency library. If sequence A aligns to B at position X, and B aligns to C at the same position, then A and C are constrained to align. This library is then used to guide the progressive alignment, producing more accurate alignments than pure greedy methods. T-Coffee is slower than progressive aligners but significantly more accurate on divergent sequences. The method can incorporate structural information through 3D-Coffee and Expresso extensions.

T-Coffee is the alignment tool that asks: what if we used logic, not just similarity, to build alignments? The answer is more accurate but more expensive — a trade-off that defines the upper frontier of MSA research.