Global Analysis of Deep Learning Prediction Using Large-Scale In-House Kinome-Wide Profiling Data

Characterization of prevalent tyrosine kinase inhibitors and their challenges in glioblastoma treatment

Glioblastoma multiforme (GBM) is a highly aggressive malignant primary tumor in the central nervous system. Despite extensive efforts in radiotherapy, chemotherapy, and neurosurgery, there remains an inadequate level of improvement in treatment outcomes. The development of large-scale genomic and proteomic analysis suggests that GBMs are characterized by transcriptional heterogeneity, which is responsible for therapy resistance. Hence, knowledge about the genetic and epigenetic heterogeneity of GBM is crucial for developing effective treatments for this aggressive form of brain cancer. Tyrosine kinases (TKs) can act as signal transducers, regulate important cellular processes like differentiation, proliferation, apoptosis and metabolism. Therefore, TK inhibitors (TKIs) have been developed to specifically target these kinases. TKIs are categorized into allosteric and non-allosteric inhibitors. Irreversible inhibitors form covalent bonds, which can lead to longer-lasting effects. However, this can also increase the risk of off-target effects and toxicity. The development of TKIs as therapeutics through computer-aided drug design (CADD) and bioinformatic techniques enhance the potential to improve patients' survival rates. Therefore, the continued exploration of TKIs as drug targets is expected to lead to even more effective and specific therapeutics in the future.

Large-Scale Modeling of Sparse Protein Kinase Activity Data

Protein kinases are a protein family that plays an important role in several complex diseases such as cancer and cardiovascular and immunological diseases. Protein kinases have conserved ATP binding sites, which when targeted can lead to similar activities of inhibitors against different kinases. This can be exploited to create multitarget drugs. On the other hand, selectivity (lack of similar activities) is desirable in order to avoid toxicity issues. There is a vast amount of protein kinase activity data in the public domain, which can be used in many different ways. Multitask machine learning models are expected to excel for these kinds of data sets because they can learn from implicit correlations between tasks (in this case activities against a variety of kinases). However, multitask modeling of sparse data poses two major challenges: (i) creating a balanced train-test split without data leakage and (ii) handling missing data. In this work, we construct a protein kinase benchmark set composed of two balanced splits without data leakage, using random and dissimilarity-driven cluster-based mechanisms, respectively. This data set can be used for benchmarking and developing protein kinase activity prediction models. Overall, the performance on the dissimilarity-driven cluster-based split is lower than on random split-based sets for all models, indicating poor generalizability of models. Nevertheless, we show that multitask deep learning models, on this very sparse data set, outperform single-task deep learning and tree-based models. Finally, we demonstrate that data imputation does not improve the performance of (multitask) models on this benchmark set.