Tutorials
A step by step guide to ranking candidates in a given region of interest can be found here and is a good place to start.
A quickstart guide in Macromedia Flash format is also available; however it might be slightly out of date (an updated version is forthcoming).
FAQ
I entered my query and pressed "Start search" but after five minutes waiting all I got was a blank page.
Your query probably timed out. Unfortunately server resources are limited so more complex queries can sometimes take longer to execute than your browser (or our web server) is willing to wait.
Try reducing the match set to fewer genes or try again at a quiet time of day.
I searched between coordinates but not all of the genes in that region turn up. What's going on?
Different builds of the human genome have different coordinate systems. Your gene might be between one set of coordinates in one build but a totally different set of coordinates in another. SUSPECTS is based on the assembly used by Ensembl - the exact build version can be found in the top right hand side of the title bar (e.g. v28 means build 28).
Why is my favourite marker X "not found in the database"?
Suspects is based on Ensembl (you can tell which release by looking at the version number at the top right of the screen). If your marker is in Ensembl, Suspects will find it. Otherwise you'll need to use chromosomal coordinates instead.
How do I interpret the XML output from the results page?
The XML output simply uses the Perl XML::Dumper module to dump out a results hash as XML. As such, the element names have no real bearing on the data contain. You'll have to use the "key" identifier in each to work that out.
If you know of a good way to fix this short of parsing the XML file and replacing text before presenting it to the user, please get in touch.
The quickstart guide says that scores are weighted according to how reliable they are - what does this mean?
Some scores are better indications that two genes are involved in the same pathway than others. For example the fact that two genes have similar expression profiles is far less of an indicator than two genes sharing a GO term. Thus each match is weighted arbitrarily and then averaged to produce the final, weighted score.
Can I see the source code / can I run Suspects locally?
Yes, Suspects is made available under an open source licence. As it relies on helper databases the procedures for setting it up might be relatively complex, however - please get in touch to discuss your requirements.
Contact
Please email Euan Adie if you need any help with the system.