Question

How does a biological organism reacts to different temperature and light conditions?

Challenges

Development and 3d-printing of parts of the acquisition device. Automatisation of acquisition and analysis of images.

Type of data

- Image data (static and dynamic)

- 3D models

Methods

Data acquisition will be possible thanks to the open source project OpenLabFrame. One potential model organism to study is the «blob».

The first part of the project will consist in modifying the device for an adapted acquisition based on the model organism (depending on the students this part could involve Computer Aided Design and the use of 3D printers).

The second part of the project will consist in developing a protocol to automate the change of heat and light conditions and synchronize these changes with data acquisition. This part is done in Python.

Finally, the third part will focus on the analysis of the dynamic of the acquired images. This part is done in Python.

Question

What is the behaviour of a series of ribosomes running on a transcript according to various transcript structures ?

Challenges

The ribosome is a molecular complex whose function is to convert a transcript (mRNA) into a protein (molecule composed of a chain of amino acids). On a transcript, several ribosomes can be present at the same time, running through the transcript at varying speeds, reacting differently when they encounter structural elements of the transcript, and colliding with each other. These dynamics generate emergent behaviours that influence protein production.

In this project, we are interested in a particular class of structural elements of transcripts called uORF (detailed will be provided later). We have a database indicating the position of 60,000 uORFs on a wide variety of transcripts and want to study the impact of these uORFs on protein production.

For this purpose, we have developed an agent based model, allowing to numerically simulate the dynamics of ribosomes. This model allows us to observe the emergent properties of the dynamic behavior of ribosomes when the model parameters are varied. The simulations are performed using a python code while the statistical analyses of these simulations are performed using a code in R language.

During the project, the students will learn how to run the simulations and how to use the analysis code. They will be in charge of proposing hypothesis, run the corresponding simulations and analyse the result to make interpretations. Students will need to collaborate to manage different tasks together.

This project is a follow-up of a project started during the thesis of Sébastien Choteau (CENTURI thesis), co-directed by Christine Brun (TAGC) and Philippe Pierre (CIML).

Type of data

• Complete description of the composition of 60,000 uORF in human and their position on
  the associated transcripts.
• Simulation code written in Object Oriented Python
• Simulation analysis code written in R

Methods

• Object Oriented Python
• R language
• Mathematical modeling
• Simulation will be run on HPC (no prior knowledge required)

Question

Can replayed activity help memory formation of heterogeneous virtual environments?

Challenges

Wayfinding in a familiar environment is essential to our daily life, whether travelling from home to work or simply walking to the coffee machine from our office. Even if navigation aids (e.g. GPS) have developed tremendously in the last few years, the alteration of this capacity as observed in Alzheimer's disease or senescence remains very disabling. Navigation abilities rely on a set of interconnected structures in the temporal lobe notably the hippocampus. In rodents, the activity of spatially selective hippocampal principal cells, called place cells, has been discovered as animals randomly searched for food in small environments. Each of these cells increase their activity when the animal is in a specific location in its environment called its “place field”. Individual place cells have different place fields and at the population level, multiple place cells map the whole environment. Importantly, ensemble place cells activity can be used to decode the position of an animal very precisely. It is generally accepted that population of place cells form the neuronal basis of a cognitive map: an internal representation of the environment used to guide navigation. During resting periods, the sequential activation of place cells is replayed but temporally compressed during bouts of synchronized activity in the hippocampus. The aim of this project is to detect significant replay events and characterize the replayed trajectory in mice navigating virtual reality environments. Virtual reality is a useful tool to control the cues used by animals to self-locate. of particular relevance will be to decipher if replay events include cells activated by movement related information, external cues or a mixture of both.

Type of data

Data from neuronal activity recordings in mice exploring real environements from public repository where replay sequences have been isolated and characterized will be used to test and compare different methods for replay sequence detection. One’s an efficient method identified, it will be tested on data obtained in mice navigating virtual reality environments.

Methods

Explained variance, Bayesian decoding, Principal component analysis, GLM, template matching

Question

How does neural population dynamics relate to behaviorally-relevant computations?

Challenges

At any given instant, hundreds of billions of cells in our brains are lighting up in a complicated yet highly coordinated manner to give rise to our thoughts, percepts, and movements. A single neuron may be connected to thousands of other cells, sending out and receiving information through electrical impulses called spikes. From an engineering perspective, these spikes form a signal that may be viewed as a series of ones and zeros rapidly unfolding in time. Altogether, these signals reflect the ongoing computations taking place inside the nervous system, and as such, constitute a window into the brain’s inner workings. Recent advances in recording techniques have allowed experimenters to collect data from hundreds to thousands of neurons simultaneously while animals perform simple tasks. Dealing with such high-dimensional data poses important technical challenges that require elaborate methods for data mining and analysis. In this project, students will deal with datasets of increasing complexity and develop a set of analyses to extract meaningful information from the data.

Type of data

Data that will be shared by the teaching staff, under the BIDS standard data organization which is currently being extended to electrophysiology data by the members of the INT and the CONECT team. The data we will use consists of:

• publicly available recordings from the dorsomedial frontal cortex of NHPs performing a time-interval reproduction task (https://github.com/jazlab/Meirhaeghe2021)
• publicly available recordings from the motor cortex (M1/PMd) during an instructed reach-to-grasp task (https://www.nature.com/articles/sdata201855, available at the following URL in BIDS: https://gin.g-node.org/sprenger/multielectrode_grasp/src/bep_animalephys)
• https://www.biorxiv.org/content/10.1101/2021.03.30.437692v4 : V1, gratings-like, natural stimulations : extracellular electrophy recordings in cat V1

Methods

Data visualisation, neural decoding, principal component analysis, kinematic and geometric analyses of neural trajectories in high-dimensional space, hypothesis-testing, null distributions and statistics