Plain-English Explanation: covid-sanity

During the early days of the COVID-19 pandemic, thousands of research papers were being published on preprint servers like bioRxiv and medRxiv—often before peer review. This project solves a real problem: it helps researchers and interested people cut through the noise and find the papers that actually matter to them. Instead of scrolling through raw feeds, you get a searchable, sortable database that lets you discover papers by topic and find similar ones with a single click.

The project works by downloading COVID-19 papers from bioRxiv and medRxiv, then using a machine-learning technique to understand what each paper is about based on its abstract. When you search for a paper or ask "what's similar to this one?", the system compares abstracts to find matches. It's like having someone who's read all the papers help you find the ones relevant to your question. The website, hosted at biomed-sanity.com, presents all this in a clean interface where you can browse, sort, and search papers easily.

Who uses this? Researchers, public health officials, medical students, and informed journalists during a crisis when the volume of new information is overwhelming. Instead of spending hours finding one relevant paper, you can find twenty in minutes. The creator also added an optional feature that pulls Twitter discussion for each paper, so you can see what people are saying about it online.

The project is built as a simple web application using Flask (a Python framework for websites). It's designed to be easy to run on your own computer for testing, but also straightforward to deploy to the internet using standard hosting tools. The creator even included their exact setup—running an hourly update script to keep the paper database fresh and showing how they host it cheaply on Linode. This spirit of openness and practical simplicity is intentional: they mention this project follows a similar one they built for arXiv papers.