RESUMEN
Tissue-to-blood partition coefficients (Ptb) are key parameters for assessing toxicokinetics of xenobiotics in organisms, yet their experimental data were lacking. Experimental methods for measuring Ptb values are inefficient, underscoring the urgent need for prediction models. However, most existing models failed to fully exploit Ptb data from diverse sources, and their applicability domain (AD) was limited. The current study developed a multimodal model capable of processing and integrating textual (categorical features) and numerical information (molecular descriptors/fingerprints) to simultaneously predict Ptb values across various species, tissues, blood matrices, and measurement methods. Artificial neural network algorithms with embedding layers were used for the multimodal modeling. The corresponding unimodal models were developed for comparison. Results showed that the multimodal model outperformed unimodal models. To enhance the reliability of the model, a method considering categorical features, weighted molecular similarity density, and weighted inconsistency in molecular activities of structure-activity landscapes was used to characterize the AD. The model constrained by the AD exhibited better prediction accuracy for the validation set, with the determination coefficient, root mean-square error, and mean absolute error being 0.843, 0.276, and 0.213 log units, respectively. The multimodal model coupled with the AD characterization can serve as an efficient tool for internal exposure assessment of chemicals.
Asunto(s)
Peces , Relación Estructura-Actividad Cuantitativa , Animales , Reproducibilidad de los Resultados , Mamíferos , Redes Neurales de la ComputaciónRESUMEN
As a newly discovered protein posttranslational modification, lysine lactylation (Kla) plays a pivotal role in various cellular processes. High throughput mass spectrometry is the primary approach for the detection of Kla sites. However, experimental approaches for identifying Kla sites are often time-consuming and labor-intensive when compared to computational methods. Therefore, it is desirable to develop a powerful tool for identifying Kla sites. For this purpose, we presented the first computational framework termed as DeepKla for Kla sites prediction in rice by combining supervised embedding layer, convolutional neural network, bidirectional gated recurrent units, and attention mechanism layer. Comprehensive experiment results demonstrated the excellent predictive power and robustness of DeepKla. Based on the proposed model, a web-server called DeepKla was established and is freely accessible at http://lin-group.cn/server/DeepKla. The source code of DeepKla is freely available at the repository https://github.com/linDing-group/DeepKla.