An uncertainty-based interpretable deep learning framework for predicting breast cancer outcome.

BACKGROUND: Predicting outcome of breast cancer is important for selecting appropriate treatments and prolonging the survival periods of patients. Recently, different deep learning-based methods have been carefully designed for cancer outcome prediction. However, the application of these methods is still challenged by interpretability. In this study, we proposed a novel multitask deep neural network called UISNet to predict the outcome of breast cancer. The UISNet is able to interpret the importance of features for the prediction model via an uncertainty-based integrated gradients algorithm. UISNet improved the prediction by introducing prior biological pathway knowledge and utilizing patient heterogeneity information. RESULTS: The model was tested in seven public datasets of breast cancer, and showed better performance (average C-index = 0.691) than the state-of-the-art methods (average C-index = 0.650, ranged from 0.619 to 0.677). Importantly, the UISNet identified 20 genes as associated with breast cancer, among which 11 have been proven to be associated with breast cancer by previous studies, and others are novel findings of this study. CONCLUSIONS: Our proposed method is accurate and robust in predicting breast cancer outcomes, and it is an effective way to identify breast cancer-associated genes. The method codes are available at: https://github.com/chh171/UISNet .

n-Butylidenephthalide Modulates Autophagy to Ameliorate Neuropathological Progress of Spinocerebellar Ataxia Type 3 through mTOR Pathway.

Spinocerebellar ataxia type 3 (SCA3), a hereditary and lethal neurodegenerative disease, is attributed to the abnormal accumulation of undegradable polyglutamine (polyQ), which is encoded by mutated ataxin-3 gene (ATXN3). The toxic fragments processed from mutant ATXN3 can induce neuronal death, leading to the muscular incoordination of the human body. Some treatment strategies of SCA3 are preferentially focused on depleting the abnormal aggregates, which led to the discovery of small molecule n-butylidenephthalide (n-BP). n-BP-promoted autophagy protected the loss of Purkinje cell in the cerebellum that regulates the network associated with motor functions. We report that the n-BP treatment may be effective in treating SCA3 disease. n-BP treatment led to the depletion of mutant ATXN3 with the expanded polyQ chain and the toxic fragments resulting in increased metabolic activity and alleviated atrophy of SCA3 murine cerebellum. Furthermore, n-BP treated animal and HEK-293GFP-ATXN3-84Q cell models could consistently show the depletion of aggregates through mTOR inhibition. With its unique mechanism, the two autophagic inhibitors Bafilomycin A1 and wortmannin could halt the n-BP-induced elimination of aggregates. Collectively, n-BP shows promising results for the treatment of SCA3.