Network biology approaches to understanding life
With living systems, it’s all about networks. How are we developing tools to understand life on earth, and sharing them openly with the community?
How can systems biology approaches help us to understand the interactions between living systems? From ecosystems to microbiomes, the information we can glean is important for improving human health and better understanding the evolution of host-specificity of bacteria, among many other things.
Márton Olbei, a PhD student in the Korcsmáros lab at EI, has taken to his PhD with flying colours, already contributing to two book chapters published in the last few months, referring to SalmoNet and SignaLink - computational tools that allow scientists to better understand the networks that underpin biology: from entire systems through to complex signalling networks.
We asked Márton, whose latest book chapter was published in Foodborne Bacterial Pathogens, to fill us in on these very useful pieces of software, and how they are helping us to better understand life from a network-based approach:
Importantly, too, how we are sharing the protocols for developing such tools, enabling the wider life science community to use and embrace this way of working.
Hi! It’s going great thanks! I’m currently on a bit of a break doing my PIPS. It’s a three month long internship students in the DTP PhD programme go through. The point is to get some professional experience that’s different from what they go through during their PhD studies.
I’m doing mine on building research culture and community at the NBI, as part of the Courage wellbeing project. It’s very different compared to what I do normally, it’s more akin to social sciences and psychology. I am currently reading up on a lot of literature. I gained a lot of respect for researchers in these fields, doing these studies is not a small task.
I’d like to establish myself as a capable researcher and communicator, someone who can do meaningful scientific work. To be more specific, I’d like to be comfortable with carrying out various computational analyses and planning experiments or projects in general.
I would like to learn to better gather my thoughts and write them up in a manner that can be published, and is beneficial to the larger scientific community (and the general public as well, I’ve always been interested in that).
I would also like to learn a bit about the less science-y part of science, about getting funding, applying for grants, organizing events and building communities.
My PhD project is about studying the evolution of host adaptation in Salmonella. Some pathogens are generalists, capable of infecting many hosts, while others are specialists only focusing on a few or one specific host species (most pathogenic bacteria fall into the first category). I am studying what molecular interactions are unique to these specialist pathogens, if there are recurring mechanisms, or if they are the results of independent adaptation.
Host adaptation is a process that is often coupled with genome degradation, and horizontal gene transfer, but in my opinion, it is not purely a genomics problem. It is the way those absent or newly acquired genes fit into the system that eventually lead the changes making Salmonella alter its behaviour within the host organism, which makes a system level analysis warranted.
The benefit of multi-layered systems level approaches is that since we are studying differences in molecular interactions, they hint at the mechanisms of change, and they also extend the scope of what we look at.
As an example, let’s say we see a transcription factor - target gene interaction missing in one of our interesting pathogens. From this, we can not only guess (and validate) the cause of that (mutation in a transcription factor binding site), but can also look at the further implications of that missing interaction (a change in regulation changing behaviour).
These resources act as collections of knowledge from multiple sources, which is useful on its own - especially when it comes to non-model organisms. Since they contain interaction data they are in a way more than the sum of their parts.
Using multiple layers (e.g. protein-protein interactions, transcription factor - target gene interactions, metabolic interactions) gives a more complete picture of the inner workings of the biological system, which is useful for all interested communities, be they network biologists like us, or molecular biologists interested in how, let’s say, a specific pathway fits into the larger scope of the organism they are studying.
SalmoNet can be utilised by a lot of researchers with different backgrounds. It’s a resource of molecular interactions of Salmonella, so understanding and discovering new biological mechanisms or potential therapeutic intervention targets is one of the most compelling uses.
The great thing about it is that the network backbone, the blueprint of the Salmonella strains can be made context specific by integrating multi-omics data on top of it, further extending the dimension of uses. The standardized, well documented formats the database uses make this a fairly simple task.
The main difference between the tools, is that SignaLink is aimed at collating interaction data of human and model organisms (the worm C. elegans, the fly Drosophila and the fish D. rerio) while SalmoNet contains data from bacterial pathogens, in particular host adapted and generalist Salmonella serovars.
However, the structure, methodology and philosophy behind the resources are quite similar.
The aim of these protocols is to show how network resources like SalmoNet and SignaLink are made.
By making the workflow accessible we can share with other research groups that building such a resource is not only feasible, but also a huge benefit to their respective communities, especially if they are working with lesser known organisms.
Provided you have the necessary information, a database like this can empower and help novel research going into these areas.
My dream is that we can push the current SalmoNet pipeline to a level where (provided we have all the information we need) we can go really quickly “from genome to network”, where a computational step like that can become routine.
Just as an example, imagine the possibilities and results we could get out of the 10K Salmonella genomes project - if we had not just the 10 SalmoNet networks, but 10,000 Salmonella networks to analyse.
Right now, I’m really excited about a big upcoming Salmonella conference, the 'Salmonella Biology and Pathogenesis' Gordon Research Conference, in June. I was lucky enough to be selected to go, and I really want to impress there.
During our first year as PhD students we have to write a probation report and literature review from our first couple of months, and it seems like most of the people I cited are going to be there...
Also, I’ve never been to the US, and it’s been somewhat of a childhood dream of mine to visit someday.
I would definitely like to stay in research and continue working with networks in some fashion. I think the beauty of this field is that you are not tied to a biological subject, and can apply your systems level expertise to a lot of questions.
(Marton is a student funded through a BBSRC NRP DTP grant number BB/M011216/1.)