Prediction and recommendation by machine learning through repetitive internal validation for hepatic veno-occlusive disease/sinusoidal obstruction syndrome and early death after allogeneic hematopoietic cell transplantation.

Using traditional statistical methods, we previously analyzed the risk factors and treatment outcomes of veno-occlusive disease/sinusoidal obstruction syndrome (VOD/SOS) after allogeneic hematopoietic cell transplantation. Within the same cohort, we applied machine learning to create prediction and recommendation models. We analyzed 2572 transplants using eXtreme Gradient Boosting (XGBoost) to predict post-transplant VOD/SOS and early death. Using the XGBoost and SHapley Additive exPlanations (SHAP), we found influential factors and devised recommendation models, which were internally verified by repetitive ten-fold cross-validation. SHAP values suggested that gender, busulfan dosage, age, forced expiratory volume, and Disease Risk Index were significant factors for VOD/SOS. The areas under the receiver operating characteristic curves and the areas under the precision-recall curve of the models were 0.740, 0.144 for all VOD/SOS, 0.793, 0.793 for severe to very severe VOD/SOS, and 0.746, 0.304 for early death. According to our single feature recommendation, following the busulfan dosage was the most effective for preventing VOD/SOS. The recommendation method for six adjustable feature sets was also validated, and a subgroup corresponding to five to six features showed significant preventive power for VOD/SOS and early death. Our personalized treatment set recommendation showed reproducibility in repetitive internal validation, but large external cohorts should prospectively validate our model.

Patient-Level Cancer Prediction Models From a Nationwide Patient Cohort: Model Development and Validation.

BACKGROUND: Nationwide population-based cohorts provide a new opportunity to build automated risk prediction models at the patient level, and claim data are one of the more useful resources to this end. To avoid unnecessary diagnostic intervention after cancer screening tests, patient-level prediction models should be developed. OBJECTIVE: We aimed to develop cancer prediction models using nationwide claim databases with machine learning algorithms, which are explainable and easily applicable in real-world environments. METHODS: As source data, we used the Korean National Insurance System Database. Every Korean in ≥40 years old undergoes a national health checkup every 2 years. We gathered all variables from the database including demographic information, basic laboratory values, anthropometric values, and previous medical history. We applied conventional logistic regression methods, light gradient boosting methods, neural networks, survival analysis, and one-class embedding classifier methods to effectively analyze high dimension data based on deep learning-based anomaly detection. Performance was measured with area under the curve and area under precision recall curve. We validated our models externally with a health checkup database from a tertiary hospital. RESULTS: The one-class embedding classifier model received the highest area under the curve scores with values of 0.868, 0.849, 0.798, 0.746, 0.800, 0.749, and 0.790 for liver, lung, colorectal, pancreatic, gastric, breast, and cervical cancers, respectively. For area under precision recall curve, the light gradient boosting models had the highest score with values of 0.383, 0.401, 0.387, 0.300, 0.385, 0.357, and 0.296 for liver, lung, colorectal, pancreatic, gastric, breast, and cervical cancers, respectively. CONCLUSIONS: Our results show that it is possible to easily develop applicable cancer prediction models with nationwide claim data using machine learning. The 7 models showed acceptable performances and explainability, and thus can be distributed easily in real-world environments.

PTK6 inhibits down-regulation of EGF receptor through phosphorylation of ARAP1.

PTK6 (also known as Brk) is a non-receptor-tyrosine kinase containing SH3, SH2, and catalytic domains, that is expressed in more than 60% of breast carcinomas but not in normal mammary tissues. To analyze PTK6-interacting proteins, we have expressed Flag-tagged PTK6 in HEK293 cells and performed co-immunoprecipitation assays with Flag antibody-conjugated agarose. A 164-kDa protein in the precipitated fraction was identified as ARAP1 (also known as centaurin delta-2) by MALDI-TOF mass analysis. ARAP1 associated with PTK6 in an EGF/EGF receptor (EGFR)-dependent manner. In addition, the SH2 domain of PTK6, particularly the Arg(105) residue that contacts the phosphate group of the tyrosine residue, was essential for the association. Moreover, PTK6 phosphorylated residue Tyr(231) in the N-terminal domain of ARAP1. Expression of ARAP1, but not of the Y231F mutant, inhibited the down-regulation of EGFR in HEK293 cells expressing PTK6. Silencing of endogenous PTK6 expression in breast carcinoma cells decreased EGFR levels. These results demonstrate that PTK6 enhances EGFR signaling by inhibition of EGFR down-regulation through phosphorylation of ARAP1 in breast cancer cells.

Molecular dissection of the interaction between the SH3 domain and the SH2-Kinase Linker region in PTK6.

PTK6 (also known as Brk) is an intracellular tyrosine kinase that contains SH3, SH2, and tyrosine kinase catalytic (Kinase) domains. The SH3 domain of PTK6 interacts with the N-terminal half of the linker (Linker) region between the SH2 and Kinase domains. Site-directed mutagenesis and surface plasmon resonance studies showed that a tryptophan residue (Trp44) in the SH3 domain and proline residues in the Linker region, in the order of Pro177, Pro175, and Pro179, contribute to the interaction. The three-dimensional modeled structure of the SH3-Linker complex was in agreement with the biochemical data. Disruption of the intramolecular interaction between the SH3 domain and the Linker region by mutation of Trp44, Pro175, Pro177, and Pro179 markedly increased the catalytic activity of PTK6 in HEK 293 cells. These results demonstrate that Trp44 in the SH3 domain and Pro177, Pro175, and Pro179 in the N-terminal half of the Linker region play important roles in the SH3-Linker interaction to maintain the protein in an inactive conformation along with the phosphorylated Tyr447-SH2 interaction.