Enhanced tumor cell killing by ultrasound after microtubule depolymerization.

Recent studies show that tumor cells are vulnerable to mechanical stresses and undergo calcium-dependent apoptosis (mechanoptosis) with mechanical perturbation by low-frequency ultrasound alone. To determine if tumor cells are particularly sensitive to mechanical stress in certain phases of the cell cycle, inhibitors of the cell-cycle phases are tested for effects on mechanoptosis. Most inhibitors show no significant effect, but inhibitors of mitosis that cause microtubule depolymerization increase the mechanoptosis. Surprisingly, ultrasound treatment also disrupts microtubules independent of inhibitors in tumor cells but not in normal cells. Ultrasound causes calcium entry through mechanosensitive Piezo1 channels that disrupts microtubules via calpain protease activation. Myosin IIA contractility is required for ultrasound-mediated mechanoptosis and microtubule disruption enhances myosin IIA contractility through activation of GEF-H1 and RhoA pathway. Further, ultrasound promotes contractility-dependent Piezo1 expression and localization to the peripheral adhesions where activated Piezo1 allows calcium entry to continue feedback loop. Thus, the synergistic action of ultrasound and nanomolar concentrations of microtubule depolymerizing agents can enhance tumor therapies.

The Effectiveness of Multitask Learning for Phenotyping with Electronic Health Records Data.

Electronic phenotyping is the task of ascertaining whether an individual has a medical condition of interest by analyzing their medical record and is foundational in clinical informatics. Increasingly, electronic phenotyping is performed via supervised learning. We investigate the effectiveness of multitask learning for phenotyping using electronic health records (EHR) data. Multitask learning aims to improve model performance on a target task by jointly learning additional auxiliary tasks and has been used in disparate areas of machine learning. However, its utility when applied to EHR data has not been established, and prior work suggests that its benefits are inconsistent. We present experiments that elucidate when multitask learning with neural nets improves performance for phenotyping using EHR data relative to neural nets trained for a single phenotype and to well-tuned baselines. We find that multitask neural nets consistently outperform single-task neural nets for rare phenotypes but underperform for relatively more common phenotypes. The effect size increases as more auxiliary tasks are added. Moreover, multitask learning reduces the sensitivity of neural nets to hyperparameter settings for rare phenotypes. Last, we quantify phenotype complexity and find that neural nets trained with or without multitask learning do not improve on simple baselines unless the phenotypes are sufficiently complex.