Not really my own work or anything but since I'll be hosting the servers and adminy things like that I thought it would be worthwhile to post up here to advertise the service. Some of the guys in the lab are interested in docking and finding targets for anti-malarial drugs. They have taken a grid based approach and have developed a BOINC client to help solve the problem of identifying new compounds. Their goal is to discover new drugs and they will need users to sign up to do some of the computation on their own computers in the @home way.

Good luck to them and please do get in touch if you are interested - FMAH .

The Scientist has published its Top Places to Work and Drug Monkey has pointed out that Universities don't feature well in the list. Instead its the private and public research institutes that feature highly in the rankings.

Given the lack of such institutes in Ireland I wonder how well the Universities here rank in terms of how good a place they are to work. Note that the survey is strictly about research capacity, rather it also ranks childcare availability and such things as part of measures to access work/life balance for researchers in these institutes. How would Irish and European Institutes and Universities fare in the same test? I reckon it would probably be a very mixed bag.

I was having a debate with a colleague about the merits of requiring or just allowing perhaps users to receive results via e-mail. Personally I've never seen the point of receiving results by e-mail, and definitely never seen the point of a website requiring it.

The disorder predictors available offer a case study in the effectiveness of such measures.
IUPred (no restriction) 246
Pondr (restricted) and incidentally the best 94
Disopred (restricted webserver but allows access to DB of preprocessed results) 157
RONN (no restriction) 188
Globplot (no restriction) 380
Disembl (no restriction) 396

I think this shows pretty nicely makes the point that requiring an e-mail is bad for citations.

I recently attended a new course run by SFI at the Ryan Academy on training scientists to setup new businesses. Whilst the majority of the interest is inspin outs it did get me thinking about business ideas and how to translate research expertise into viable companies. Having completed a number of projects with commercial interests I do have a decent insight into how the world of industry works. However, this has normally been in the context of much larger well established firms. Typically projects are of a collaborative nature where the company does not necessarily see any conflict between the work they are asking the researcher to do and their core business. Ireland at the moment is in a good position to expand its biotech and software development business. Despite all the bad news in the press many large companies are still moving jobs here. The problem for Ireland is that these are normally foreign firms without a vested interest in the country and therefore the long term stability of employment is still under question since these firms can easily relocate. The challenge therefore for Irish Researchers is how to generate an indigenous research and development industry. What I learned from this course is that perhaps now is the time to be trying to do just that if you are a young researcher. The money seems to be available, wages are low after all the pay cuts everyone has been taking and the opportunities are still there. Hopefully, we will see a number of firms starting up capitalising on the large pool of talented Irish researchers who will now have an opportunity to work in Ireland. There is a new call for proposals under the scheme. Applicants should contact their Technology Transfer Offices and start planning the next stage of their commercialisation process.

Round up of the course from DCU's Gordon ?McConnell.

Together with Aidan Budd and Toby Gibson at the EMBL Heidelberg I will be organising an EMBO Practical Course on Protein Bioinformatics Tools at the Advanced Training Centre in Heidelberg in September. The course will teach wet-lab biologists the latest tools to identify functional modules in their protein of interest and introduce them to concepts like protein disorder, modular architecture and linear motifs.

• How can I come up with new ideas for the function of my proteins of interest?
• What is already known about the function of these proteins?

This course presents bioinformatics tools and concepts that can help address these two questions that are crucial for the work of many experimental/wet-lab biologists. Wet-lab researchers, with limited experience using such tools, yet who often need to answer these questions in their research, are the main target group of the course.

In particular, the course will cover analyses of protein sequences, 3D structures, interactions, and networks. Additionally, we will focus on proteins as modular entities, with both intrinsically ordered and disordered modules responsible for specific aspects of protein function.

Trainers include developers of several key tools for analysis of this kind, including Pfam, IUPRED, STRING, Scan Site, ELM, and PhosphoELM. Some of the other tools and resources presented will include Jal View, BLAST, REFLECT, SMART, ?UniProt, PyMOL, PDB, and STITCH. Most tools used during the course are web-based and can be used, free of charge, by academic scientists worldwide.

To apply for the course, fill in the online application form linked to from this site http://www.embl.de/training/events/2011/AIN11-01/registration/index.html

More information is available at http://bioinfo-casl.ucd.ie/empa

I noticed an interesting paper this week that details a new experimental technique applied to human cell lines. The method is capable of measuring the half life of proteins in human cell lines for the first time. We have been looking at Yeast data from the O'Shea lab. The yeast data is pretty comprehensive and all available for download. The yeast work uses an epitope tagged strain. Their work in yeast suggests that protein half life correlates well with transcriptional regulation (in some protein clusters).

The new work in humans uses a method called bleach-chase to achieve quantitative measurements of YFP tagged proteins. Importantly they are able to distinguish between proteins that undergo active degradation and those that are duiluted due to the effects of cell growth. That is when the cell divides no more of that particular protein is produced resulting in an overall lower abundance of the protein in the new daughter cells.

They were able to show that proteins with a long half life are generally regulated through dilution whereas those with short protein half lives are typically regulated by protease activity.

A new version of SLiMSearch is available on the Bioware server. This version allows searching for motifs against the human genome. We are looking to increase the coverage of species - get in touch if you want your model organism added. Currently we prefer to work with ?UniProt proteomes, because of the quality of the Gene Ontology Annotation.

We have a testing version of a new piece of software from the labs. Called cyclops the tool is designed to help in the generation of large numbers of constrained and linear peptides. Non-natural amino acids and other commercially available residues from the ZINC database can also be incorporated.

The program is capable of switching between different representations of the peptide, for example the SMILES description. The output of the program give empirical information on peptide properties advising the user on the likely ease of solid-phase peptide synthesis, and provides options to filter the library on the basis of druglike properties of the peptides.

It is available on Bioware.

The students on the Structured PhD Program in bioinformatics have organised a Symposium to highlight their work and bring together scientists working on innovative work. It will be held on December 6th and 7th with some excellent keynote speakers. It is open to all and promises to be an interesting couple of days. The full program is available online.

Had a few discussions in the lab today where we required an ontology or set of Gene Ontology terms to describe a protein complex rather than its constituents. I've seen ontologies for constituent parts of the cell but not for more transient complexes. In part the need for this should be guided by the likes of the systems biology graphical notation framework where the idea of units of activity come together to form distinct activities in cooperation. However, I have still haven't seen an actual ontology of protein complexes.