Forum:BSc Computer Science dissertation project ideas

Question

[cars4urkids]

Hi,

I'm currently studying computer science and I wish to specialise in bio-computing, I need to work on a project and write a dissertation for my degree and I was wondering if I could get some help in narrowing down which topics would be realistic. I have a basic background in biology and I'm interested in epigenetics so I wish to somehow incorporate that into my project. A software engineering based project would be preferred.

My current idea is to make a (visual?) tool based on Bayesian networks to predict likelihood of disease or expression of a gene associated with a disease based on already known links with e.g. high levels of methylation. Firstly is there much data available about these sorts of links? Is there a need for a tool like this? Is it feasible to estimate the likelihood of disease or at least suggest genes to target for study?

My second idea is also to create a tool to make a Boolean network model of gene regulatory networks and use methylation data to influence the GNR. Would there be a useful application of this in the real world?

I know very little details about both topics so I'm hoping this post hasn't just been a string of key words but actually makes sense!

Thank you in advance!

Answer 1

[Aerval]

I am not sure I can truly answer this as I am a student my self as I only have the following advice: Try to get in contact with a local group that works in bioinformatics (or some other scientific computing, the topic of your bachelor thesis is less important than you may think) and work with them. You may think that you can pull off something nice on your own with just a good topic but that (in all honesty) should not be what you want. Although you may be able to learn most things and ask all your questions on the internet, there is nothing that will help you more than having a local advisor who you can ask when you have problems. I once had an advisor whom I saw once every two weeks but even that level of contact was enough to reflect on my own work and see what I have to do next and where I may be totally wrong.

[DG]

I second this. No other easier way to do work that is completely irrelevant than failing to at least talk with local experts in the field. But coming to Biostars is also a good start!

Answer 2

[DG]

Discussing with your supervisor is of course a must, I am assuming this is essentially what elsewhere we would call an Honours Project or Senior Project? So you want your scope to be fairly tight and not overly broad if you only have about 6 months to a year to work on it and write it up. Finding local experts to discuss things with would also be helpful so you can avoid doing what's already been done. I'm not an expert on either of the two things you propose, but i suspect there has already been work in the area. A lot depends on how much time you have to do this work and what the requirements for novelty are.