Random Forest Team
ProBAI-Guided ProB Evolution to Ease Feedback InhibitionCandidate Analysis, Strain Construction and HPLC Peak Comparisons
ProBAI-Guided ProB Evolution to Ease Feedback Inhibition
Random Forest Team studies E. coli ProB sites D107 and E143 through strain construction, HPLC peak-area comparisons and computational analysis.
- Team name
- Random Forest Team
- Institution
- School of Biological Science and Technology, University of Jinan
Team media is for event showcases and reporting only. Portrait rights remain with the individuals. For concerns about portrait use, contact the organizers. We will investigate and take appropriate action.
Team Motto
“Decoding Biology for Productive Cells ”
Background
Feedback Limits in a Cell Factory
Research Approach
Experiments First, Analysis Next
- Enzyme and Sites
- ProB · D107、E143
- Construction Strategy
- Wild Type · Single · Double Mutants
- Experimental Readout
- HPLC Main-Peak Area
The first design round focuses on ProB residues D107 and E143.
Parallel construction and controls compare the effects of site combinations.
Peak areas compare samples; they are not presented as unverified absolute product yields.
Research Process
Development & Validation
- 01
Stage 1
Three Screening Layers
The planned workflow scans over 7,000 mutations across ProB's 367 positions with ESM-2, refines candidates using DynaMut2 stability and AutoDock Vina binding estimates, then validates D107N, E143A and the double mutant. In the actual first round, wet-lab work preceded computational interpretation. - 02
Stage 2
Preparing HPLC Samples
The first derivatization and HPLC preparation took nearly a day: culture extraction, centrifugation, reagent and mobile-phase preparation, sonication, filtration, tubing and pump flushing. Prepared assay samples were the concrete output of this stage. - 03
Stage 3
Peak Areas and a Weak Model Baseline
Some mutant groups show higher HPLC peak areas than wild type. Random-forest leave-one-out R² is −0.3551, indicating performance below prediction with the training mean. Peak areas are in mAU·min and are not absolute yields.
Project Evidence
Methods & Figures
Teamwork
Computational, experimental and candidate review records
Project Stories
Project & Team Records
Computational Screening and HPLC Checks
Team films
