Search

Can we develop an interpretable machine learning model that optimizes starting gonadotrophin dose selection in terms of mature oocytes (metaphase II [MII]), fertilized oocytes (2 pronuclear [2PN]) and usable blastocysts?This was a retrospective study of patients undergoing autologous IVF cycles from 2014 to 2020 (n = 18,591) in three assisted reproductive technology centres in the USA. For each patient cycle, an individual dose-response curve was generated from the 100 most similar patients identified using a K-nearest neighbours model. Patients were labelled as dose-responsive if their dose-response curve showed a region that maximized MII oocytes, and flat-responsive otherwise.Analysis of the dose-response curves showed that 30% of cycles were dose-responsive and 64% were flat-responsive. After propensity score matching, patients in the dose-responsive group who received an optimal starting dose of FSH had on average 1.5 more MII oocytes, 1.2 more 2PN embryos and 0.6 more usable blastocysts using 10 IU less of starting FSH and 195 IU less of total FSH compared with patients given non-optimal doses. In the flat-responsive group, patients who received a low starting dose of FSH had on average 0.3 more MII oocytes, 0.3 more 2PN embryos and 0.2 more usable blastocysts using 149 IU less of starting FSH and 1375 IU less of total FSH compared with patients with a high starting dose.This study demonstrates retrospectively that using a machine learning model for selecting starting FSH can achieve optimal laboratory outcomes while reducing the amount of starting and total FSH used.

The increasing integration of software and automation in modern chemical laboratories prompts special emphasis on two important skills in the chemistry classroom. First, students need to learn the technical skills involved in modern scientific computing and automation. Second, applying these techniques in practice requires effective collaboration in teams. This work aims at developing a teaching module to help students gain both skills. In particular, we describe a modular and collaborative approach for introducing undergraduate students to scientific computing in the context of automated and autonomous chemical laboratories. Using online collaboration tools, students work in parallel teams to develop central components of an automated computer vision system that monitors color changes in ongoing chemical reactions. These components include three different aspects: image capture, communication, and data visualization. The image capture team collects and stores the images of the chemical reaction, the communication team processes the images, and the visualization team develops the tools for analyzing the processed image data. Using this educational framework, students built an open-source Python tool called AutoVis that enables the automated tracking of color and intensity changes in a liquid. The software is tested by simulating chemical reactions with dilute solutions of food coloring in water. It is shown that the system reliably tracks color and intensity, providing feedback to the experimentalist and enabling further computational analysis. Over the course of the project, students gain proficiency in scientific computing using Python and collaborate on software development using GitHub. In this way, they learn the role of software in chemical laboratories of the future.

A drastic transition at birth, from constant maternal nutrient supply in utero to intermittent postnatal feeding, requires changes in the metabolic system of the neonate. Despite their central role in metabolic homeostasis, little is known about how pancreatic β cells adjust to the new nutritional challenge. Here, we find that after birth β cell function shifts from amino acid- to glucose-stimulated insulin secretion in correlation with the change in the nutritional environment. This adaptation is mediated by a transition in nutrient sensitivity of the mTORC1 pathway, which leads to intermittent mTORC1 activity. Disrupting nutrient sensitivity of mTORC1 in mature β cells reverts insulin secretion to a functionally immature state. Finally, manipulating nutrient sensitivity of mTORC1 in stem cell-derived β cells in vitro strongly enhances their glucose-responsive insulin secretion. These results reveal a mechanism by which nutrients regulate β cell function, thereby enabling a metabolic adaptation for the newborn.