University of South Dakota Showcase

Research hypothesis: Vehicles with larger engines (higher displacement and horsepower) and greater weight tend to have lower fuel efficiency (mpg). Dataset: The dataset contains information on several automobile models. Each row represents a unique vehicle, and the dataset includes key technical and performance variables such as: name, mpg, cylinders, displacement, horsepower, weight, acceleration, model_year, and origin. Use the data for: Exploratory Data Analysis (EDA), regression modeling, teaching, etc. To inteprete the dataset: Compare variables, look for patterns, trends, etc. Limitation: Small sample dataset, the source could not be independently verified.

Urban wastewater microbiomes are complex and temporally dynamic, offering valuable insight into community-scale microbial ecology and potential public health trends. However, existing wastewater-based studies often remain descriptive, lacking tools for predictive modeling. In this study, we introduce a digital twin framework that forecasts microbial abundance trajectories in urban wastewater using an interpretable generative model, Q-net. Trained on a 30-week longitudinal metagenomic dataset from seven wastewater treatment plants, the model captures temporal microbial dynamics with high fidelity (R² > 0.97 for key taxa; R² = 0.998 at the final timepoint). Beyond accurate forecasting, Q-net provides transparent model structure through conditional inference trees and enables simulation of realistic microbial trends under hypothetical scenarios. This work demonstrates the potential of digital twins to move wastewater microbiome studies from static snapshots to dynamic, predictive systems, with broad implications for environmental monitoring and microbial ecosystem modeling. More detail in https://www.biorxiv.org/content/10.1101/2025.07.21.666059v1

A wide range of chemicals have been found to cause negative chemotaxis in Escherichia coli. Here, we report Penicillin-Streptomycin (Pen-Strep) as a repellent that can repel different strains of E. coli. E. coli DH5α grown in LB broth. These bacterial solutions were exposed to 4% agar loaded with PenStrep as a drug releasing system in 12-well plates. Penicillin at 2500 units/ml (ug/ml for Streptomycin) concentrations (25%) of the PenStrep-Agar loading system were prepared for chemotaxis evaluation. Repelling of the E. coli was clearly observed via microscopy within 30 mins of the exposure. Moreover, continuous video acquisition and analysis by real-time recording over 8 hours showed a distinct region with a high density of cells in the region away from the repellant and vice-versa for the region close to the repellant. The optical density measurements indicated a considerable reduction in cell aggregation near the repellent, while a majority of the cells were observed to be distributed farther away from the repellent area. Stimulated single cell trajectory tracking showed movement away from the repellant in the y-direction at a velocity of 0.017 µm/s when exposed to 25% PenStrep chemical gradient. Unstimulated cells showed normal motility with repeated run-and-tumble movement in random direction in the control groups. This study provides a proof of concept for high-throughput analysis of the selective chemotaxis behavior of the microbial community by microbiome engineering to achieve the goal of beneficial biofilm formation to inhibit biofilm formation by corrosive bacteria.

The challenges faced with bioscience lab protocols encourage the development of automation systems to come alongside technicians. Today’s experiments involve complex protocols with precise actions usually performed manually. Even simpler biologicals can be tedious and error-prone, as was seen during the COVID-19 pandemic and society’s demand for high-volume, rapid analysis. Moreover, reproducibility suffers from excessive variability and insufficient data, while low-income labs remain slow to adopt automation. To approach these challenges and more, we leveraged aspects of the Stanford Biodesign process to develop a modular lab automation system with open-source components. Our system’s modularity minimizes the need for multiple inflexible systems, instead allowing for removing individual components as needed. Modularizing the software architecture enables us to insert artificial intelligence and machine learning models, bringing greater value to our data. Soon, using open-source chatbots such as ChatGPT will assist in converting manual protocol steps to automated functions at a low cost. Combining all of these available, low-cost tools can ultimately bring more democratization and widespread adoption of bioscience automation.

Experimental studies exploring the effects of intranasal oxytocin are typically underpowered due to small samples. Open access to experimental data and procedures and the use of previously employed measures is critical to building more robust and replicable findings, especially in less studied areas of oxytocin research. In this paper, data is provided from a double-blind placebo-controlled crossover study exploring the effects of intranasal oxytocin (IN-OT: 24 IU) on social preference to romantic partners, parents, peers, and strangers. Young adults (N=44; 91% female) in committed dating relationships completed three phases of data collection including a screening survey followed by two laboratory visits. In addition to romantic partner-, and stranger attraction ratings, the data is the first to provide comparisons between attachment and social preference ratings to parents, close friends, and romantic partners under placebo and IN-OT conditions. The data also include differences by situational and life history factors known to moderate oxytocin effects. The detailed protocol, and dataflow can be accessed to verify the analysis and findings or to conduct a replication study. The standardized experimental design and common IN-OT protocol add to the capacity for a meta-analysis exploring oxytocin effects on partner preference, and may also be directly ported to existing or future studies with related questions to increase sample size and power.

This grip strength data was collected for older adults registered to compete in the National Senior Games between 2011 and 2017. Grip was taken with a hydraulic hand grip dynamometer set in position 2. Athletes were positioned with their arm by their side and their elbow bent to 90 degrees. Grip was collected in alternating fashion between the right and left hands for a total of three tirlas on each hand. Dominant hand was reported.