A systematic evaluation of the performance of GPT-4 and PaLM2 to diagnose comorbidities in MIMIC-IV patients.

Background: Given the strikingly high diagnostic error rate in hospitals, and the recent development of Large Language Models (LLMs), we set out to measure the diagnostic sensitivity of two popular LLMs: GPT-4 and PaLM2. Small-scale studies to evaluate the diagnostic ability of LLMs have shown promising results, with GPT-4 demonstrating high accuracy in diagnosing test cases. However, larger evaluations on real electronic patient data are needed to provide more reliable estimates.

A Modelling Framework Linking Resource-Based Stochastic Translation to the Optimal Design of Synthetic Constructs.

The effect of gene expression burden on engineered cells has motivated the use of "whole-cell models" (WCMs) that use shared cellular resources to predict how unnatural gene expression affects cell growth. A common problem with many WCMs is their inability to capture translation in sufficient detail to consider the impact of ribosomal queue formation on mRNA transcripts. To address this, we have built a "stochastic cell calculator" (StoCellAtor) that combines a modified TASEP with a stochastic implementation of an existing WCM.