SSA:Using machine learning to improve data analysis from complex in vivo datasets:lifespan cellular resolution images of the zebrafish musculoskeletal
Lead Research Organisation:
University of Bristol
Department Name: Physiology and Pharmacology
Abstract
Full title: Using machine learning to improve data analysis from complex in vivo datasets: lifespan cellular resolution images of the zebrafish musculoskeletal and healthy joint ageing
The skeletal system is surprisingly dynamic, undergoing remodelling due to changes in gene expression
and or loading throughout life. Changes to either underpin skeletal and joint diseases and over 15
million people in the UK have a musculoskeletal disorder such as osteoporosis or osteoarthritis.
Zebrafish are increasingly used as the animal model of choice to study developmental biology and cell
behaviour. They offer excellent genetic tractability along with dynamic in vivo imaging due to their
translucency and the potential to use fluorescent reporters to track cells in the whole animal. Our
group has made >20 mutant lines of zebrafish carrying mutations in genes that lead to disease states
in humans, along with transgenic lines that allow us to see the cells that make up muscle, cartilage,
bone, tendons and the immune system in living fish. We have amassed a large number of 3D datasets
that contain data that we currently do not fully extract. This project focuses on developing machine
learning strategies to process large, complex, 3D in vivo datasets with the aim of using these to
develop high throughput systems for testing of new clinically relevant genes and in vivo compound
screening to test new pharmaceutical strategies. The project is highly interdisciplinary offering the
chance to combine advanced in vivo skills (CRISPR genome editing, live imaging of transgenic
reporters) with computational AI and machine learning approaches to visualise and analyse data. The
supervisory team has members from both academia and industry. The project would give the student
a highly desirable skill set that is increasingly in huge demand. This project would particularly suit a
student with an interest in biological systems and some experience of programming, ideally in Python.
The skeletal system is surprisingly dynamic, undergoing remodelling due to changes in gene expression
and or loading throughout life. Changes to either underpin skeletal and joint diseases and over 15
million people in the UK have a musculoskeletal disorder such as osteoporosis or osteoarthritis.
Zebrafish are increasingly used as the animal model of choice to study developmental biology and cell
behaviour. They offer excellent genetic tractability along with dynamic in vivo imaging due to their
translucency and the potential to use fluorescent reporters to track cells in the whole animal. Our
group has made >20 mutant lines of zebrafish carrying mutations in genes that lead to disease states
in humans, along with transgenic lines that allow us to see the cells that make up muscle, cartilage,
bone, tendons and the immune system in living fish. We have amassed a large number of 3D datasets
that contain data that we currently do not fully extract. This project focuses on developing machine
learning strategies to process large, complex, 3D in vivo datasets with the aim of using these to
develop high throughput systems for testing of new clinically relevant genes and in vivo compound
screening to test new pharmaceutical strategies. The project is highly interdisciplinary offering the
chance to combine advanced in vivo skills (CRISPR genome editing, live imaging of transgenic
reporters) with computational AI and machine learning approaches to visualise and analyse data. The
supervisory team has members from both academia and industry. The project would give the student
a highly desirable skill set that is increasingly in huge demand. This project would particularly suit a
student with an interest in biological systems and some experience of programming, ideally in Python.
Organisations
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| BB/R505626/1 | 01/10/2017 | 30/09/2021 | |||
| 2117425 | Studentship | BB/R505626/1 | 24/09/2018 | 02/04/2023 | Abdelwahab Kawafi |
| Description | This project focuses on improving the analysis of CT scans and 3D microscopy images. So far I have developed two machine learning models, one to automatically detect bones and cartilage from CT scans, and one to detect and track particles and cells from microscopy. |
| Exploitation Route | New zebrafish lines are constantly being generated with bone disease mutations and shared between labs, more are constantly being CT scanned which this model can be used for automated analysus |
| Sectors | Digital/Communication/Information Technologies (including Software),Healthcare,Pharmaceuticals and Medical Biotechnology |
| URL | https://github.com/wahabk/ctfishpy |