Software Daily Summary
Software Daily Summary: What Happened on November 8, 2026
Open-source tool predicts drug targets with ChEMBL 37
A separate guide describes how researcher Stephen Turner built personal and lab websites with Markdown and open-source tools.
Today in brief
A new open-source tool applies the Similarity Ensemble Approach to ChEMBL 37, with a Python API and web interface for predicting potential drug targets.1 In a suvorexant test, its five top-ranked targets were similar to, but not identical to, those from the original implementation.1
Stephen Turner described using Markdown, Pandoc, GitHub and Netlify for personal and lab websites, including a 5 KB lab site that loads in about 300 milliseconds.2
Open-source target-prediction tool uses ChEMBL 37
The open-source tool implements the Similarity Ensemble Approach (SEA), which predicts potential drug targets by comparing chemical similarity, using ChEMBL 37.1 Its Python API and web interface support SMILES, a text format for chemical structures, and the interface displays a two-dimensional structure.1
In a suvorexant test, the tool's five highest-ranked targets were similar to those from the original implementation, but the lists did not match exactly.1 The author suggested that differences in dataset versions and parameters may explain the results. The tool uses ChEMBL 37, while the original site uses ChEMBL 36.1 The author plans to refine the code and test it with other ligands.1
Turner builds personal and lab websites with lightweight tools
Stephen Turner builds his personal website from Markdown files and a Makefile, then deploys it through a private GitHub repository and Netlify.2 He said the site is about 100 KB and that its domain costs $8 a year.2
Turner said he launched a lab website the previous week, using Make and Pandoc to generate its HTML.2 The site is 5 KB and loads in about 300 milliseconds, he said.2 Its publications page uses a BibTeX file and Pandoc Lua filters to format citations and emphasize lab members' names.2