Forum:Seeking ideas and guidance for an undergraduate project

Question

[silverbullet]

This is my first post on here. I am a 3rd year undergraduate student doing a degree in medical physiology. For my third year I took a unit for computational biology as well as a bioinformatics project.

This is my first time ever dealing with maths or computer science in the realm of biology and I would appreciate some guidance regarding my project. I a project that can be realistically achieved in one academic year by someone with no previous experience. I am only beginning to learn python but coding is not mandatory for the project.

I have read about about a few options:

1. Structure prediction (homology modelling/ab initio/threading)
   • Can this be realistically achieved?
   • Would I need to use python?
   • How would I go about finding an interesting protein that no one has modeled yet?
2. Small drug discovery
   • Can this be realistically achieved?
   • Would I need to use python?
   • Do I need to know much about chemistry?
3. Neural networks
   • What sort of projects involve neural networks and can I realistically achieve these?

Please if you have any other ideas, I am eager to hear them out. Any projects that can be done by someone with no prior experience but passion and an understanding of biology. I don't like genetics so much but I understand that they are primary for a big part of bioinformatics.

Any extra information regarding the steps I could take would be very helpful. And I hope that this post isn't typical and annoying in this community.

[Pierre Lindenbaum]

Undergraduate Project topic

need ideas for bioinformatics research project (undergraduates)

Project For A Beginner Bioinformatics Student

...

[steve]

You can replicate some of the aspects of the study here: https://genomemedicine.biomedcentral.com/articles/10.1186/s13073-017-0424-2

You can download some data from here: https://portal.gdc.cancer.gov/repository

I just did one of these for a tumor burden analysis and it was not too hard and pretty fun. Get the 'masked somatic mutation' .maf files and go to town.

Programming skills needed are minimal, you can probably get a decent amount of it done in Excel.

Answer 1

[Carlo Yague]

Structure prediction (homology modelling/ab initio/threading)

This is realistic. I have done something similar when I was undergraduate (my project was to design mutations that stabilizing two of the four structural configurations of a protein). However, its probably too ambitious to develop your own structure prediction tool. Instead, you should focus on using some of the many tools available and carefully interpret the results of your predictions. If you want to use python, modeller, a tool for homology modelling, is used within python. Make also sure to visualize your structures with pyMol. You don't need to know that much about chemistry to get your project running. I would say that the minimum is to understand the different classes of amino acids (polar, charged, ...) and the concept of free energy.