PLK1 as a cooperating partner for BCL2-mediated antiapoptotic program in leukemia

Abstract

The deregulation of BCL2 family proteins plays a crucial role in leukemia development. Therefore, pharmacological inhibition of this family of proteins is becoming a prevalent treatment method. However, due to the emergence of primary and acquired resistance, efficacy is compromised in clinical or preclinical settings. We developed a drug sensitivity prediction model utilizing a deep tabular learning algorithm for the assessment of venetoclax sensitivity in T-cell acute lymphoblastic leukemia (T-ALL) patient samples. Through analysis of predicted venetoclax-sensitive and resistant samples, PLK1 was identified as a cooperating partner for the BCL2-mediated antiapoptotic program. This finding was substantiated by additional data obtained through phosphoproteomics and high-throughput kinase screening. Concurrent treatment using venetoclax with PLK1-specific inhibitors and PLK1 knockdown demonstrated a greater therapeutic effect on T-ALL cell lines, patient-derived xenografts, and engrafted mice compared with using each treatment separately. Mechanistically, the attenuation of PLK1 enhanced BCL2 inhibitor sensitivity through upregulation of BCL2L13 and PMAIP1 expression. Collectively, these findings underscore the dependency of T-ALL on PLK1 and postulate a plausible regulatory mechanism.

AlphaML: A clear, legible, explainable, transparent, and elucidative binary classification platform for tabular data

“…The core of alphaML machine learning is composed of various modules including data management, feature selection, data normalization, data sampling, hyperparameter tuning, model training and validation, testing, and model interpretability, reflecting a similar approach we have recently embraced. 7 The data management module initially verifies the data accessibility in a particular directory of the local storage. If absent, it retrieves example datasets from FigShare.…”

“…The rationale behind the construction of NegLog2RMSL has been elucidated in prior work. 7 In brief, when we assess the performance of a model, using multiple performance metrics is beneficial as it offers a holistic evaluation, encompassing various facets of prediction quality. A single metric might only provide a snapshot of the model’s capability, but integrating several metrics paints a more detailed picture.…”

“…The methodology employed for hyperparameter tuning incorporated a compound scoring system that utilized both Cohen’s kappa coefficient and the Matthew’s correlation coefficient (MCC). 7 This amalgamated scoring strategy ensures a rigorous evaluation of categorization models by harnessing the unique capabilities inherent in each metric. Cohen’s kappa, with its numerical values spanning from −1 to 1, quantifies the concurrence between two independent classifiers, effectively controlling for random chance and imbalance in class distribution.…”

“…This method was also described previously. 7 Alongside this, the "kappa_mcc_error" was utilized in conjunction with a controlled fitting approach, implemented to mitigate both overfitting and underfitting tendencies in our model. We used a "custom_score", which incorporates the "kappa_mcc_error" in a bipartite manner, taking into account both training and validation scores.…”

“…Recent research endeavors utilizing disease-centric and drug-specific biological attributes have exhibited notable predictive efficacy. 4 , 5 , 6 , 7 , 8 Nevertheless, these methodological approaches necessitate substantial investment in terms of time and resources and demand a synergetic collaboration between researchers in the fields of biology and data science to ensure the effectiveness and rationality of the strategy.…”