%0 Journal Article %T Probabilistic Sequence Alignments: Realistic Models with Efficient Algorithms %+ Bioinformatique Structurale %A Yeramian, Edouard %A Debonneuil, Edouard %Z The work was supported by the Ministère de la Recherche Scientifique (ACI-IMPBIO-2004 GENEPHYS) and the Institut Pasteur (DVPI contract and Strategic Horizontal Programme on Anopheles gambiae). %< avec comité de lecture %@ 0031-9007 %J Physical Review Letters %I American Physical Society %V 98 %N 7 %P 078101 %8 2007-02-16 %D 2007 %R 10.1103/PhysRevLett.98.078101 %M 17359063 %Z PACS numbers 87.10.+e, 05.10.-a, 87.15.Cc %Z Life Sciences [q-bio]/Quantitative Methods [q-bio.QM]Journal articles %X Alignment algorithms usually rely on simplified models of gaps for computational efficiency. Based on correspondences between alignments and structural models for nucleic acids, and using methods from statistical mechanics, we show that alignments with realistic laws for gaps can be computed with fast algorithms. Improved performances of probabilistic alignments with realistic models of gaps are illustrated. By contrast with optimization-based alignments, such improvements with realistic laws are not observed. General perspectives for biological and physical modelings are mentioned. %G English %Z We thank Michael Nilges for constant scientific support and Richard Miles and Olivier Martin for careful reading of the manuscript. %2 https://pasteur.hal.science/pasteur-01570174/document %2 https://pasteur.hal.science/pasteur-01570174/file/71%20YERAMIAN.pdf %L pasteur-01570174 %U https://pasteur.hal.science/pasteur-01570174 %~ PASTEUR %~ CNRS