As a contributor to Faculty of 1000, we ask that you highlight papers that interest you. To assist you, we offer to search for recent papers that might be of interest. We will construct search profiles according to your specifications to suggest articles that you may want to evaluate.

You can create as many profiles as you like; there are three different types of profile searches that you can create:

1. Cluster Search Profiles (more details below): which we highly recommend as they will provide the best results: you supply us with model articles related to a theme of interest, and we then use a specially developed search tool that takes advantage of the PubMed clustering algorithm to find related papers.
2. Author Name Profiles: you enter the names of authors you are interested in.
3. Boolean (Keyword) Profiles: you enter keywords, author names and / or phrases relating to areas of research that interest you.
   
   

To set up a profile, go to My Details and click on Add/modify profile.

Why should I bother with Cluster searches?
Most faculty members who have set up cluster searches (or had us set them up) are delighted with the results.

In general, Cluster searches provide more relevant "hits" and less noise than Key word searches, unless searching for papers on a really simple theme (eg "Parkinsons disease" or "cyclins AND rat"); however, as soon as you start searching for a longer string of key words, a Cluster search tends to provide better results.

What is a Cluster search?
A Cluster search is a "smart" literature search: it uses a group of sample papers on a theme to identify relevant papers of interest. The program works optimally when you're able to supply 10-15 sample papers related to your theme of interest.

Sounds great but I don't have time to set it up
We'll do it for you - either contact the staff member you deal with regularly or e-mail editorial@f1000biology.com with the theme(s) you want covered (and, if you have time, some sample papers).

DIY Guide to setting up your first Cluster profile: Set up a separate profile for each distinct theme

1. Go to http:///my/profiles/cluster on the Contributor site, click on the green PubMed link mid way down the page (or go to http://www.ncbi.nlm.nih.gov/pubmed/ ).
2. Author Name Profiles: you enter the names of authors you are interested in.
3. In PubMed, do a search to identify sample papers relevant to a theme of interest (e.g. "transplants and Parkinsons") and copy/cut and paste the 8-digit PubMed ID (PMID) number -- the last line of each paper citation -- into the box indicated on the contributor site page for each paper you want included in your Cluster profile.
4. Once you've entered all the paper PMIDs for a theme (in general it's a good idea to have 10 or more), then click on the "Continue" arrow at the bottom. You will be shown a full detailed list of the paper's you've selected; then click on "Continue" to see a Preview page listing the types of papers that your cluster profile will provide (nb the Preview page estimate of how many results there will be each week can sometimes be misleading so best to try out for "real" for one week).
5. Check that results listed on the preview page match your theme. The preview profile results page is generally biased towards the very most relevant papers so if you notice "noise" at this stage, then the amount of noise in the alerts will be probably be higher.
6. If happy with the results, click on "Submit", if not, then click on "Modify" and see "Trouble-shooting" section below.

Trouble-shooting

1. We're happy to help
   Our editorial staff will be happy to help (either the person you deal with regularly or e-mail editorial@f1000biology.com). It will be easier for them to help you if you can tell them what's missing or the type of noise (eg "my search is returning loads of cancer papers that are nothing to do with my theme") or if you can identify particular articles that the search should or should not pick up.
   
   
2. DIY: troubleshooting yourself
   (i) Too few papers
   Try adding more sample papers to the cluster profile. It's usually a good idea to have at least 10-15 sample papers per cluster profile. Adding more sample papers to a cluster profile will increase the number of results (it may also increase the noise a bit). A cluster profile of three sample papers on a clearly identifiable theme can often be enough to identify the most relevant papers on that theme, but may not pick up many less obvious papers.
   
   (ii) Too much noise although some relevant papers
   Try examining the sample papers you used for the cluster profile and then remove likely papers that are misleading the program. In general mixing more than one theme in a cluster search increases the noise, and sometimes completely throws the results. Adding even one misfitting paper on a theme to a cluster profile will increase the noise, sometimes dramatically.
   
   (iii) Papers seem irrelevant
   Try examining the results for possible mixed themes that have confused the search (eg sample papers on "transcription regulation" mixed with papers on "T4 signalling" may well produce only results on "transcription regulation by T4") and then separate the two themes into two Cluster profiles. It's fine to include reviews as sample papers but best to stick to specific titles (dementia in Parkinsonism) and avoid vague whimsical titles (eg "wither wanders the brain") which will confuse the program utterly.