Carregant...
Fitxers
Tipus de document
TesiVersió
Versió publicadaData de publicació
Tots els drets reservats
Si us plau utilitzeu sempre aquest identificador per citar o enllaçar aquest document: https://hdl.handle.net/2445/193945
Desarrollo de herramientas para el análisis y predicción patogénica de las variantes missense de ATM en el entorno clínico
Títol de la revista
Autors
Director/Tutor
ISSN de la revista
Títol del volum
Recurs relacionat
Resum
[spa] Establecer la naturaleza patogénica de las variantes de secuencia en ATM, un gen asociado con el cáncer de mama y otros cánceres hereditarios, es crucial para brindar una atención adecuada a los pacientes. Sin embargo, lograr buenas clasificaciones de estas variantes sigue siendo un problema sin resolver. Aquí, abordamos este problema mejorando la contribución de las herramientas in silico a la clasificación de variantes missense.
Un problema importante en el uso de herramientas in silico es su baja interpretabilidad. Nos acercamos a esta limitación explorando primero el desarrollo de predictores de patogenicidad específicos de proteínas que incorporan medidas interpretables del impacto de una variante. Paralelamente, desarrollamos una familia de representaciones gráficas rápidas e intuitivas en las que se considera el impacto de una variante en relación al de variantes patogénicas y benignas ya conocidas. Entre los resultados obtenidos destaca un sistema de clasificación para el criterio de predicción in silico de la adaptación de las guías ACMG/AMP a ATM y tres predictores específicos de proteína con diferentes grados de interpretabilidad. Dos tienen capacidades predictivas similares a las de los predictores de patogenicidad de alto rango. Curiosamente, aunque menos preciso, nuestro predictor
biofísico alcanza un rendimiento que abre el camino para usar evidencia biofísica para completar la anotación de variantes de ATM.
En lo que respecta a las representaciones gráficas, las hemos aplicado a tres problemas de clasificación de variantes: evaluación de predicciones, caracterización de VUS y comparación de las dos versiones de las guías ACMG/AMP adaptadas a ATM. En estas aplicaciones, nuestros gráficos muestran sus virtudes como herramientas complementarias para los procesos de clasificación in silico: son rápidos, fáciles de usar y casi no requieren capacitación.
En resumen, en esta tesis presentamos una familia de herramientas in silico para mejorar la anotación de variantes missense en ATM y facilitar el papel de los profesionales en este proceso.
[eng] Establishing the pathogenic nature of sequence variants in ATM, a gene associated with breast cancer and other hereditary cancers, is crucial to providing adequate patient care. However, achieving good rankings for these variants remains an unresolved issue. Here, we address this problem by improving the contribution of in silico tools to missense variant classification. A major problem in the use of in silico tools is their low interpretability. We approach this limitation by first exploring the development of protein-specific pathogenicity predictors that incorporate interpretable measures of the impact of a variant. In parallel, we developed a family of fast and intuitive graphical representations in which the impact of a variant is considered in relation to that of known pathogenic and benign variants. Among the results obtained, a classification system stands out for the in silico prediction criterion of the adaptation of the ACMG/AMP guides to ATM and three protein-specific predictors with different degrees of interpretability. Two have predictive abilities similar to those of high-rank pathogenicity predictors. Interestingly, although less accurate, our biophysical predictor achieves a performance that paves the way for using biophysical evidence to complete ATM variant annotation. Regarding the graphical representations, we have applied them to three variant classification problems: prediction evaluation, VUS characterization and comparison of the two versions of the ATM-adapted ACMG/AMP guides. In these applications, our graphs show their virtues as complementary tools for in silico classification processes: they are fast, easy to use and require almost no training. In summary, in this thesis we present a family of in silico tools to improve the annotation of missense variants in ATM and facilitate the role of professionals in this process.
[eng] Establishing the pathogenic nature of sequence variants in ATM, a gene associated with breast cancer and other hereditary cancers, is crucial to providing adequate patient care. However, achieving good rankings for these variants remains an unresolved issue. Here, we address this problem by improving the contribution of in silico tools to missense variant classification. A major problem in the use of in silico tools is their low interpretability. We approach this limitation by first exploring the development of protein-specific pathogenicity predictors that incorporate interpretable measures of the impact of a variant. In parallel, we developed a family of fast and intuitive graphical representations in which the impact of a variant is considered in relation to that of known pathogenic and benign variants. Among the results obtained, a classification system stands out for the in silico prediction criterion of the adaptation of the ACMG/AMP guides to ATM and three protein-specific predictors with different degrees of interpretability. Two have predictive abilities similar to those of high-rank pathogenicity predictors. Interestingly, although less accurate, our biophysical predictor achieves a performance that paves the way for using biophysical evidence to complete ATM variant annotation. Regarding the graphical representations, we have applied them to three variant classification problems: prediction evaluation, VUS characterization and comparison of the two versions of the ATM-adapted ACMG/AMP guides. In these applications, our graphs show their virtues as complementary tools for in silico classification processes: they are fast, easy to use and require almost no training. In summary, in this thesis we present a family of in silico tools to improve the annotation of missense variants in ATM and facilitate the role of professionals in this process.
Matèries (anglès)
Citació
Col·leccions
Citació
PORRAS MONROY, Luz marina. Desarrollo de herramientas para el análisis y predicción patogénica de las variantes missense de ATM en el entorno clínico. [consulta: 28 de gener de 2026]. [Disponible a: https://hdl.handle.net/2445/193945]