Completed & Discontinued Projects
2024
Quantifying the drivers of genetic change in plant breeding
Summary: Plant breeding programmes are a complex network of a multitude of operations and decisions. Therefore quantifying which operations and decisions drive genetic change in plant breeding is challenging. Traditionally we measure the genetic change with a phenotypic or genetic trend, but these trends measure only change in the mean. To understand genetic change and its drivers more comprehensively, we also need to measure changes in genetic mean and variance and evaluate drivers of changes in mean and variance.
2023
Developing high performance black soldier fly breeds for insects-as-feed sector
Summary: The rising demand for livestock and aquaculture protein necessitates sustainable alternatives to soy and fishmeal, with the black soldier fly emerging as a key solution. This project aims to develop a breeding program and tools to enhance the genetics and production efficiency of black soldier fly, supporting the growing number of global producers.
A hierarchical approach for evaluating athlete performance
Summary: The hierarchical approach for evaluating athlete performance utilizes mixed-effects regression models and principal component analysis to create the ON score, a comprehensive metric that captures an athlete’s overall contribution to their team. This method, validated on NBA data, offers a reliable and efficient tool for coaches and managers to assess and compare athlete performance across seasons and games, enhancing decision-making in sports analytics by providing deeper insights into individual and team dynamics.
Publications:
A comparative study of markerless and marker-based systems
Summary: This project aims to evaluate the agreement between markerless and marker-based motion capture systems using 95% functional limits of agreement within a linear mixed-effects modeling framework. By comparing these systems, the study seeks to determine the reliability and applicability of markerless technology in biomechanical analysis, which could significantly streamline data collection and reduce setup time while maintaining accuracy. This research will provide valuable insights for improving motion capture methods used in sports and clinical biomechanics.
Publications:
2022
AlphaPart - Partition of Breeding Values by Paths of Information
Summary: The partitioning method is described in Garcia-Cortes et al. (2008). The package includes the main function AlphaPart for partitioning breeding values and auxiliary functions for manipulating data and summarizing, visualizing, and saving results.
Publications:
R Package:
Genomic strategies for optimal crossbreeding in African dairy cattle
Summary: Develop genomic strategies to optimise crossbreeding in livestock breeding programmes with focus on East African crossbred dairy cattle
2021
The use of biostatistics for optimizing athletes performance
Summary: Uniquely blending Data Science and Sports Science to generate customized strategies by athlete
Publications:
Development of predictive models and analytics techniques to forecast historical data-driven outcomes
Summary: Predictive modelling is a commonly used statistical technique to predict future behavior. Predictive modelling solutions are a form of data-mining technology that analyses historical and current data and generates a model to help predict future outcomes. In predictive modelling, data is collected, a statistical model is formulated, predictions are made, and the model is validated as additional data becomes available.
Publications:
- Global short-term forecasting of COVID-19 cases, 2021
- Modelling menstrual cycle length in athletes using state-space models, 2021
Optimising selection and management of diversity in plant breeding
Summary: This project aims to quantify and manage genetic variation in plant breeding programs, ensuring long-term competitiveness by measuring and distributing genetic diversity across different stages and years.
Publications:
2020
The lcc Package
Summary: Longitudinal concordance correlation (LCC) is a statistical measure used to assess the agreement between two methods of measurement over time. It combines aspects of precision and accuracy to evaluate the consistency and accuracy of longitudinal data, accounting for both within-subject correlation and temporal changes. LCC is particularly useful in fields where repeated measurements are essential, such as medical research, to ensure that methods produce reliable and consistent results across different time points.
Publications:
- lcc: Longitudinal Concordance Correlation, 2019
- lcc: an R package to estimate the concordance correlation, Pearson correlation and accuracy over time, 2020
R Package:
Transmission efficiency of xylella fastidiosa
Summary: Xylella fastidiosa is genetically diverse and has many vector species. However, there is limited information on vector specificity and efficiency for different sequence types (STs) Both STs of X. fastidiosa and vectors differ in their associations with plants
Publications:
- Transmission efficiency of xylella fastidiosa subsp. Pauca sequence types by sharpshooter vectors after in vitro acquisition, 2018
- Settling and feeding behavior of sharpshooter vectors on plum genotypes with different susceptibility levels to leaf scald disease (Xylella fastidiosa), 2020
2019
Sugarcane straw management for bioenergy
Summary: Global warming can intensify the soil organic matter (SOM) turnover, damaging soil health. Crop residues left on the soil are important to maintain a positive SOM budget and nutrient cycling. But, sugarcane (Saccharum officinarum) straw has been removed from the field for bioenergy purposes. We hypothesize that global warming, together with straw removal, will negatively impact Brazil’s ethanol carbon footprint.
Publications:
Measuring color using image analysis
Summary: Promote the usage of image analysis as well as development of statistical methodologies for that purpose.
Publications: