https://hdl.handle.net/20.500.12008/46869 Thu, 23 Apr 2026 11:48:26 GMT 2026-04-23T11:48:26Z https://colibri.udelar.edu.uy:443/jspui/retrieve/239640/LOGO 1.jfif https://hdl.handle.net/20.500.12008/46869 https://hdl.handle.net/20.500.12008/51574 Título: Transformers for genomic prediction : working with Yeast and Wheat traits Autor: Castro, Graciana; Hoffman, Romina; Musitelli, Mateo; Fariello, María Inés; Lecumberry, Federico Resumen: AI is becoming state-of-the-art across scientific fields, giving novel solutions to age-old problems. In genomic prediction, Machine Learning methods could not outperform linear regressions in a general way yet, but are becoming closer. An important feature when working with genomic data, which is non other than a long sequence of information, is to account for the linkage disequilibrium, i.e. dependencies between genome variations that do not need to be close in the genome, and variate with respect to the reference genome. To explode this feature, we evaluate Transformers, known for their great performance with long sequences. We worked with two databases: the first one composed of Yeast SNPs seeking to predict the growth of each individual in two different environments and the second one composed of Wheat SNPs seeking to predict four phenotypes. We compare the results with different linear models (BRR, BayesA, BayesB, BayesC and BayesL) typically used for genomic prediction and also with XGBoost, commonly known to have well performance in the area. We conclude that Transformers have shown to be a competitive model for genomic prediction, even tho it does not achieve the state-of-the-art yet. Wed, 01 Jan 2025 00:00:00 GMT https://hdl.handle.net/20.500.12008/51574 2025-01-01T00:00:00Z https://hdl.handle.net/20.500.12008/51573 Título: Deep learning for genomic prediction and tasks learned on the way Autor: Fariello, María Inés; Arboleya, Lucía; Belzarena, Diego; Castro, Graciana; De los Santos, Leonardo; Elenter, Juan; Etchebarne, Guillermo; Hoffman, Romina; Hounie, Ignacio; Musitelli, Mateo; Lecumberry, Federico Sun, 01 Jan 2023 00:00:00 GMT https://hdl.handle.net/20.500.12008/51573 2023-01-01T00:00:00Z https://hdl.handle.net/20.500.12008/51571 Título: PredGenIA : Transformers para predicción genómica Autor: Castro, Graciana; Hoffman, Romina; Musitelli, Mateo; Fariello, María Inés; Lecumberry, Federico Resumen: Nuestro proyecto busca aplicar el algoritmo de los Transformers, en la rama de la predicción genómica. Debido a los buenos resultados obtenidos de aplicar este algoritmo en el campo del NLP, y los posibles paralelismos entre los datos genómicos y los datos lingüísticos, es que nos planteamos analizar si los mismos buenos resultados se obtienen en este nuevo campo. Mon, 01 Jan 2024 00:00:00 GMT https://hdl.handle.net/20.500.12008/51571 2024-01-01T00:00:00Z https://hdl.handle.net/20.500.12008/51365 Título: Superfluous edges and exponential expansions of De Bruijn and Kautz graphs Autor: Canale, Eduardo A.; Gómez, José Resumen: A new way to expand De Bruijn and Kautz graphs is presented. It consists of deleting superfluous sets of edges (i.e., those whose removal does not increase the diameter) and adding new vertices and new edges preserving the maximum degree and the diameter. The number of vertices added to the Kautz graph, for a fixed maximum degree greater than four, is exponential on the diameter. Tables with lower bounds for the order of superfluous sets of edges and the number of vertices that can be added, are presented. Thu, 01 Jan 2004 00:00:00 GMT https://hdl.handle.net/20.500.12008/51365 2004-01-01T00:00:00Z