Automated pipeline for processing fluorescence calcium imaging microscopy data and electrophysiological measurements from 3D brain organoids

Abstract

Motivation: 3D brain organoids are biological models of human brain development from which a rich body of data can be extracted that allows for insights about neuronal dynamics. Several tools, such as fluorescence calcium imaging microscopy and electrophysiological recordings are being used to assess how 3D brain organoid dynamics recapitulate what is known from the human brain. In-silico automated workflows that process such datasets on a large scale in a robust and reproducible manner need to be developed.
Results: An automated, modular and scalable workflow was written using Python 3.9. ImageJ, Inscopix API and Caiman/CNMFe that outputs neuronal cell traces and additional scripts were produced to enable large scale analysis of Axion multi-electrode array output.

Popular Abstract

Decoding brain clones

Cloning has not belonged to the realm of science fiction for decades, but recent advances in how to “reprogram” ordinary cells, making them act like any kind of tissue found in the human body are creating new frontiers in medical and pharmaceutical sciences.

For some people, the most lasting impression of Biology lessons in school are the dissections. Biologists, it seemed, love to dissect anything that moved. The difference between mollusk, fish and frog illustrates the evolutionary history of cardiovascular systems well, to name one thing. In this case, the animals were models that show us how these biological systems work. While to many people, this may seem a bit morbid, model organisms used in experiments have been vital to bring many of the advances in life science that we now take for granted. The word vaccine for example, is derived from the Latin word for cow.

For a long time, animals have served well as models in life science, much due the ingenuity of researchers and methods such as gene editing, but to answer some questions, we need models that are much closer to the human organ that we are trying to investigate. Enter cloned organoids, simple, organ-like structures that can be created from any human.

The human brain is unique, and importantly, involved in many diseases that may be hard to reproduce in other model organisms. Many neurological conditions have a complex genetic background, and sometimes only a few individuals carry the genes necessary to replicate the condition. Making brain models with the same genetic content as these individuals makes it possible to try different treatments on a large scale and with a low cost, promising to bring new medication for several rare diseases and insight into this vital organ that plays such a large part in defining what we are.

The caveat is that while these models can now be produced on a large scale, there are not enough humans around to manually deal with the information that comes from laboratory measurements on these models. To make the next leap, software needs to be developed that can process large amounts of information using machine vision and artificial intelligence to make sense of the data. That software has now been put in production, providing researchers with the final missing steps towards decoding the messages hidden within these brain clones. Who knows what they may find?

Master’s Degree Project in Biology/Molecular Biology/Bioinformatics 30 credits 2021
Department of Biology, Lund University

Advisor: Bruno Fontinha
A:head Bio AG

• author: Knulst, Mattis
• supervisor: Bruno Fontinha
• organization: Degree Projects in Bioinformatics
• course: BINP50 20211
• year: 2021
• type: H2 - Master's Degree (Two Years)
• subject: Biology and Life Sciences
• language: English
• id: 9075470
• date added to LUP: 2022-02-16 16:31:46
• date last changed: 2022-02-16 16:31:46