Week 5 & 6 update

You can find below the continuation of my adventures in the lab :) Should you have any questions, do not hesitate to ask them !
Week 5 & 6 update
Like

Share this post

Choose a social network to share with, or copy the URL to share elsewhere

This is a representation of how your post may appear on social media. The actual post will vary between social networks

Week 5

Thankfully, I was able to conduct my experiment successfully this week. However, I had to adapt the number of flies used; I thought that it would be reasonable to have a sample of ten flies for each group, but I had to reckon that it was unrealistic.

I have indeed only managed to record the behavior of ten flies per day, and more would have taken me all night. This realization has taught me that an experience is meant to be readjusted throughout the process and that it is not fixed in time. 

The rest of my week has been occupied by labeling important genotypes to train the DLC neural network (as explained in my Week 3 & 4 update).

As forty frames per genotype are required to improve the accuracy of software predictions, this task took me two whole days and will be one of my main tasks next week.

Week 6

First of all, I had to finish the labeling part to complete our genotype library. Once it was done, the network had to be trained by all the recordings that we had labeled.

Given that we had a lot of them and that the neural network has to predict the positions of each body part and then come back to compare with the manual labeling that was done before to correct itself, it took the rest of the week to be finished. 

During this time waiting for the neural network to be ready, I continued my main experiment on the treadmill. I indeed tested the same two groups of flies that were 20 days old. Fortunately, I did not encounter any difficulties related to the computer’s storage (which was the cause of last week’s issue). 

I also mainly focused on getting the visualization of the results part done by coding with Python.

My goal was to display graphs showing the x, y coordinates as a function of time to be able to calculate the flies' speed and then compare it according to their genotype. I also had to visualize all the pairs (x, y) that the fly's legs were able to reach since this information is a good indicator of the state of the animal's motor skills.

Coding was the most challenging part of my internship so far because I have spent several days trying to solve the same issue without knowing where it comes from. It has been particularly difficult for me to complete this task considering that I only learned how to code in Python two weeks.

Nevertheless, I am firmly convinced that everything that I have learned by struggling to code will be of great help in the future, not only for my studies but also for the data analysis that I will have to run as a neuroscientist. In addition, I have never made so much progress in such a short amount of time, and I feel like it is the biggest accomplishment of my internship.

Please sign in

If you are a registered user on Laidlaw Scholars Network, please sign in

Go to the profile of Dan-Thi Nguyen
over 1 year ago

It is great to see how your research is evolving and how you face new learnings and unexpected challenges. Looking forward to reading about your two last weeks soon!