top of page

Data Challenge #3


The data for this study titled "Are all Human Embryos Mosaic?" was provided by Lauren Kelly and Darren K Griffin from the School of Biosciences at the University of Kent. The researchers investigated the rates of aneuploidy and mosaicism in human preimplantation embryos using two methods: fluorescence in situ hybridization (FISH) and next-generation sequencing (NGS). In addition to analysing published studies, the study also utilized data provided by Cooper Genomics to explore the true rates of mosaicism and aneuploidy. The goal of the research is to examine whether all human preimplantation embryos are indeed aneuploid and mosaic, as hypothesized.


The original data and preprint can be found here:


Here are some ideas for Hackathon projects:


  1. Machine Learning Model for Diagnosing Mosaicism:

    • Develop a machine learning model that predicts the likelihood of mosaicism in human preimplantation embryos based on various factors such as testing method (FISH or NGS), chromosome pairs tested, and fixation methods. Train the model using the provided data and evaluate its accuracy in predicting mosaicism in comparison to the actual reported rates.

  2. Comparison of FISH and NGS:

    • Conduct a comprehensive analysis comparing the efficacy of FISH and NGS methodologies in detecting aneuploidy and mosaicism. Explore the advantages and disadvantages of each method and visualize the variations in mosaicism rates reported by different studies.

  3. Predictive Modelling for Aneuploidy Misdiagnosis:

    • Use machine learning algorithms to predict the likelihood of aneuploidy misdiagnosis based on different fixation methods. Analyse the impact of fixation methods on the accuracy of diagnoses and identify factors contributing to misdiagnosis.

bottom of page