hello everybody,
I have planned to create a database exclusively for Pan domains. These domains play a very important role in causing malaris. also, this domain is present in diverse organisms ranging from c.elegans to homo sapiens and they mediate carbohydrate interaction.
I have included some general information about the pan domain containg sequences like name of the prtein, length, mol.wt, taxonomy, protein and dna sequnce, function of the proteins, other domans interacting with pan domains, small molecules interacting with pan domains etc..
i want to make an annotated database of pan domains. i have planned to add phylogenetic analysis of these domains and predict the secondary structure of the domian containing proteins. what more analysis work could be added to make the database a nice one.
Thanks in advance.
Janani
pan domian database
-
carolinethorn
- Former Expert
- Posts: 393
- Joined: Tue Sep 20, 2005 2:40 pm
Hi Janani,
I am assuming that youare talking about making a website type of database that will be available to the public rather than something like a protege or access db that is for use by those who have a copy of it given to them.
I was looking around to find other example of specific domain databases and did not see very many. In general most domain information is contained within the structural resources and they have a little bit about amny different domains. It would be good to have an external resources section to link to some of the other popular structure sites such as prosite, and PDB.
I found an example of an EF Hand database that I think is an example of what not to do!
http://www-user.yokohama-cu.ac.jp/~kawa ... hands.html
This example is much too text heavy. It would have really benefitted from some pictures of structures and diagrams to represent what is contained within the text.
Having a search function so that you can text search which proteins are in yoru database would be good. It would also help if you could search by gene name or gene symbol since gene names are not always the same as the protein names, or by other synonyms. You can probably pull all of that information from NCBI.
Since the domains are important for malaria i would also include links to useful malaria websites such as PlasmoDB.
I am not sure what other analysis tools would be appropriate and it sounds like a pretty big project already so maybe concentrate on what you have so far.
Good databases link out to other places which might have additional info (like NCBI) and have clear information about the sources of the sequences and the appropriate accession numbers for different resources and the build version. Make sure the pages are visible for conventional browsers (explorer, firefox and netscape for PC, if you can get hold of a mac see if it works on a mac or let people know) and have a help section where you can list known problems or bugs. I think its nice to have a "Last updated on" tag with the date so that viewers know if something has changed since they last visited. Be careful about putting your email address on there as it will invite lots of spam bots. One way to get around this is by spelling out the address in words - "so-and so at server dot com".
Best of luck.
Caroline
I am assuming that youare talking about making a website type of database that will be available to the public rather than something like a protege or access db that is for use by those who have a copy of it given to them.
I was looking around to find other example of specific domain databases and did not see very many. In general most domain information is contained within the structural resources and they have a little bit about amny different domains. It would be good to have an external resources section to link to some of the other popular structure sites such as prosite, and PDB.
I found an example of an EF Hand database that I think is an example of what not to do!
http://www-user.yokohama-cu.ac.jp/~kawa ... hands.html
This example is much too text heavy. It would have really benefitted from some pictures of structures and diagrams to represent what is contained within the text.
Having a search function so that you can text search which proteins are in yoru database would be good. It would also help if you could search by gene name or gene symbol since gene names are not always the same as the protein names, or by other synonyms. You can probably pull all of that information from NCBI.
Since the domains are important for malaria i would also include links to useful malaria websites such as PlasmoDB.
I am not sure what other analysis tools would be appropriate and it sounds like a pretty big project already so maybe concentrate on what you have so far.
Good databases link out to other places which might have additional info (like NCBI) and have clear information about the sources of the sequences and the appropriate accession numbers for different resources and the build version. Make sure the pages are visible for conventional browsers (explorer, firefox and netscape for PC, if you can get hold of a mac see if it works on a mac or let people know) and have a help section where you can list known problems or bugs. I think its nice to have a "Last updated on" tag with the date so that viewers know if something has changed since they last visited. Be careful about putting your email address on there as it will invite lots of spam bots. One way to get around this is by spelling out the address in words - "so-and so at server dot com".
Best of luck.
Caroline
-
janani10786
- Posts: 8
- Joined: Mon Feb 12, 2007 1:04 pm