Hi,
I am kind of new to bioinformatics in general and to gene sequence search in particular. I have tried BLAST but I am not sure I can use it for my specific needs.
In general, can one use BLAST with more general queries other than just looking for a specific sequence? Can one freely search sequences using SQL type queries like in a regular DB?
In particular, I'd like to find sequences in specific gene positions (for all genes of a given organism) not according to an exact sequence alignment, but rather according to some rules, for instance where there are no G's in a given portion of genes (where the reference point is the gene transcription start site). How can I do that?
BLAST strictly speaking is not a database, although it uses one. BLAST is a local homology search algorithm, and the software to implement that search program. BLAST does one thing - it takes a query sequence and finds regions of similarity in a set of sequences (which happen to be in a database) based on alignment scores.
I know of no web site or tool to do exactly what it sounds like you wish to do. Honestly, it sounds like the kind of thing where you'd be best served to download the genomic sequence data for the species you want to work with, and doing searches locally with your own code. Maybe even dump the genomic data into your own SQL database if you want to.
Thanks a lot.
I'll do that. I thought it'll just be much simpler working on the genome which is already annotated for start and end of genes.
Without these annotations, I can search for senes depending on -10 -35 promoter hexamer identification, but then I might find sequences which are actually not beginnings of genes.
you can use NCBI mapviewer , UCSC and Ensembl genome browsers to search for specific gene and its DNA sequences
You can download both a genome's sequence and its annotation from NCBI (or ensembl or the UCSC genome browser site, depending on your needs or intended uses).
I vaguely recall a tool that allowed regexes on sequence but even that is not what you want.
Biomart on Ensembl will allow you to extract sequences around a feature (e.g. upstream of a gene) so perhaps you could run your search on the extracted sequence.
If you are prepared to code, something like Biopython can allow you to represent sequence and do all kinds of operations on it (including regexes).
I am not sure this is a very easy task, but it seems to me that after retrieving your gene sequences (whole genes or just exons?) from, say, USCS, then you would need to write your own code to find what you need (are you looking at A+T-only stretches?; how long?).
Hi, The best suggestion would be, to get linked up with the Whole UCSC Gene browser where you can get fixed of all your basic doubts! Good luck!!
One of the reasons, presumably, that you cannot do more complex type of searches of gene sequences in a ready-to-use web format so readily is the same reason that Google searches are fairly limited and don't incorporate REGEXes etc. They will tend to get computationally unfeasible very quickly if used naively and they would very quickly grind to a halt with the weight of badly thought-out queries that would invariably occur.
ENSEMBL Biomart is a great menu-driven tool for extracting information with certain conditions attached to it. You should certainly take a look at it if you don't already know it.
If that still doesn't meet your need. There is a tool to do SQL type searches in ENSEMBL. It's the ENSEMBL API. If you wish to extract large amounts of information and have very specific queries, it may be worth your while learning the relevant programming skills to use it. The ENSEMBL team put a lot of effort into providing useful teaching materials and they are quite prompt in responding to queries but using the API takes time to learn how to do it. You will need to learn some Perl if you don't already know it.
http://www.ensembl.org/info/docs/api/index.html
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics