The second MIDD student symposium will take place on September 2nd at 2:30 pm in CHEM108 (Chemistry Building). The event will give student affiliated with the MIDD the opportunity to present their work to other UWM students. We hold this event six times a year, bimonthly on the first Wednesday of the month at 2:30 pm. The presentation are 20 minutes long with 5 min Q&A. We have the following presentations:
Identification and Genomic Analysis of 2-Phenylethanol Production by Lelliottia nimipressuralis JS203 with Broad-Spectrum Potency against Economic Phytopathogens
John Joseph Srok from the Research Group of Prof. Ching-Hong Yang Department of Biological Sciences, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin 53211, United States
Colletotrichum dematium, a causal agent of anthracnose, poses an escalating threat to global agriculture, exacerbated by the emergence of resistance to conventional fungicides. In this study, we employed a “one strain–many compounds” (OSMAC) approach to screen 724 environmental bacterial isolates under varied media and temperature conditions to maximize secondary metabolite diversity. Isolate JS203, identified as Lelliottia nimipressuralis through whole-genome sequencing, demonstrated robust, broad-spectrum inhibition against C. dematium and several other major phytopathogens. Bioactivity-guided fractionation of JS203 extracts, followed by NMR and GC-MS analysis, identified the volatile aromatic alcohol 2-phenylethanol (2-PE) as the primary antifungal metabolite. Genomic analysis revealed that JS203 harbors Ehrlich pathway genes, including idpC and gpr_2, which share 100% amino acid identity with validated 2-PE biosynthetic enzymes from Enterobacter sp. CGMCC 5087. These results suggest a conserved metabolic route for 2-PE production in this understudied genus. Collectively, our findings highlight L. nimipressuralis JS203 and its production of 2-PE as a promising natural strategy for the sustainable management of anthracnose and other fungal diseases.
Automated Quantification of Lateral Flow Assays at Low Analyte Concentrations
Zabina Tasneem from the Research Group of Prof. Qingsu Cheng, Department of Biomedical Engineering, College of Engineering and Applied Sciences, University of Wisconsin-Milwaukee
Lateral Flow Assays (LFAs) are widely used for rapid diagnostics. However, interpretation of these tests remains largely qualitative and relies on human visuals, limiting the ability to detect faint signals. At low analyte concentrations, test (T) lines may be visually undetectable, leading to false-negative results. Recent studies have explored machine learning based image analysis pipelines for automated quantification of LFA images; yet, reliable detection of faint test-line signals at very low analyte concentrations remains challenging. This study presents an automated image-based framework for quantitative detection of faint T-line signals in LFAs. The proposed approach preprocesses LFA images to identify the control (C) and test (T) regions of interest (ROIs), applies geometric normalization, and extracts the C–T strip region using an ensemble-based preprocessing module. A shallow U-Net architecture is used to segment relevant signal regions, enabling pixel-level analysis of the detected control and test lines. The framework is evaluated using a publicly available COVID-19 LFA dataset spanning a concentration range of 0.074–7.4 ng, which includes low analyte levels where visual interpretation becomes unreliable. Pixel-level intensities are computed within the segmented control and test regions, and a normalized T/C ratio is calculated to reduce variations caused by illumination differences and background noise. Regression analysis demonstrates a clear concentration-dependent relationship between the measured T/C signal and analyte concentration. Log–log regression yields R² values of 0.76 for average intensity and 0.86 for total intensity (p < 0.05), indicating statistically significant signal quantification across the concentration range. The framework enables computational detection of faint T-line signals, supporting the transition of LFAs from visual, qualitative interpretation toward more reliable quantitative analysis. By reducing false-negative outcomes at low analyte concentrations, the approach has the potential to improve health monitoring, enable earlier infection detection, and enhance point-of-care diagnostics.