ProteiQ is built to turn a 300-page protein report into something people can query like a chatbot.
1. cc: Life Science Podcast focuses on The Art of Protein Engineering with Carter Mitchell and Chris Conner.
2. Chris Conner hosts the show, and Carter Mitchell is the CSO of Kemp Proteins, speaking from the Advanced Lateral Flow Conference.
3. The episode asks how Kemp Proteins uses insect expression, protein engineering, and ProteiQ to solve protein-development problems.
4. Mitchell says Kemp Proteins has existed for about 30 years and started with insect-based recombinant protein production for pharmaceutical companies and government customers.
5. He describes the company changing hands, gaining a quality management system, and then adding design elements in protein engineering, machine learning, artificial intelligence, and quantitative analytics.
6. Mitchell compares E. coli, native sources, and insect cells, then explains that baculovirus can infect insect cells through a two-genome system.
7. He says baculovirus can also be functionalized as BACMAM with VSVG on the surface to transduce mammalian cells.
8. Mitchell notes that baculovirus can carry up to about 37,000 base pairs and can avoid plasmid preparation.
9. He says insect expression required more expertise than HEK-293 workflows, which became easier through Thermo Fisher tools.
10. He says new insect cell lines such as XPSF9 have made insect expression easier again.
11. Mitchell lays out four questions for a protein project: what the protein is, what information is available, what the intended use is, and what the scale is.
12. He says the intended use changes the work depending on whether the protein is for diagnostics, vaccine, therapeutic use, or another material application.
13. He explains that engineering can remove membrane embeddings or cleavage liabilities and improve consistency and homogeneity.
14. Mitchell says regulators do not want to see a smear and that assay support and documentation help remove questions.
15. He describes ProteiQ as a system built because one person cannot hold all the structural biology, purification, and engineering experience alone.
16. He says ProteiQ now summarizes data, and the next version will let operators query a chatbot for questions like isoelectric point, secretory tags, extinction coefficients, and degradation products.
17. Mitchell says the report can be 200 to 300 pages, but the goal is to predict liabilities, fix them, and shorten turnaround time.
18. He says he wants technicians to become scientists who notice nonconforming data before QC sends an alert, and he links that to scaling the company.
19. The conversation is technical, conversational, and collaborative, with Chris Conner asking clarifying questions throughout.
20. People interested in protein expression and AI workflows will get the most from it, while listeners wanting light entertainment may skip it.
