- Home
- Genomes
- Genome Browser
- Tools
- Mirrors
- Downloads
- My Data
- Projects
- Help
- About Us
Return to FAQ Table of Contents
BLAT is an alignment tool like BLAST, but it is structured differently. On DNA, BLAT works by keeping an index of an entire genome in memory. Thus, the target database of BLAT is not a set of GenBank sequences, but instead an index derived from the assembly of the entire genome. By default, the index consists of all non-overlapping 11-mers except for those heavily involved in repeats, and it uses less than a gigabyte of RAM. This smaller size means that BLAT is far more easily mirrored than BLAST. Blat of DNA is designed to quickly find sequences of 95% and greater similarity of length 40 bases or more. It may miss more divergent or shorter sequence alignments. (The default settings and expected behavior of standalone Blat are slightly different from those on the graphical version of BLAT.)
On proteins, BLAT uses 4-mers rather than 11-mers, finding protein sequences of 80% and greater similarity to the query of length 20+ amino acids. The protein index requires slightly more than 2 gigabytes of RAM. In practice -- due to sequence divergence rates over evolutionary time -- DNA BLAT works well within humans and primates, while protein Blat continues to find good matches within terrestrial vertebrates and even earlier organisms for conserved proteins. Within humans, protein Blat gives a much better picture of gene families (paralogs) than DNA Blat. However, BLAST and psi-BLAST at NCBI can find much more remote matches.
From a practical standpoint, BLAT has several advantages over BLAST:
BLAT is commonly used to look up the location of a sequence in the genome or determine the exon structure of an mRNA, but expert users can run large batch jobs and make internal parameter sensitivity changes by installing command-line Blat on their own Linux server.
First, check if you are using the correct version of the genome. For example, two versions of the human genome are currently in wide use (hg19 and hg38) and your sequence may be only in one of them. Many published articles do not specify the assembly version so trying both may be necessary.
Very short sequences that go over a splice site in a cDNA sequence can't be found, as they are not in the genome. qPCR primers are a typical example. For these cases, try using In-Silico PCR and selecting a gene set as the target. In general, the In-Silico PCR tool is more sensitive and should be preferred for pairs of primers.
Another problematic case is searching for sequences in repeats or transposons. BLAT skips the most repetitive parts of the query and limits the number of matches it finds, leading to missing matches for these repeat sequences. The online version of BLAT masks 11mers from the query that occur more than 1024 times in the genome and limits results to 16 matches per chromosome strand. This means that at most 32 locations per chromosome are returned. This is done to improve speed, but can result in missed hits when you are searching for sequences in repeats.
Often for repeat sequences, you can use the self-chain track to find the other matches, but only if the other matches are long and specific enough. You can check whether any sequence is present at a particular location by using the "Short match" track if your sequence is less than 30 bp. You can work around this minimum length limitation by adding more flanking sequence to your query to make the query unique enough. If this is not possible, the only alternative is to download the executables of BLAT and the .2bit file of a genome to your own machine and use BLAT on the command line. See Downloading BLAT source and documentation for more information. When using the command line version of BLAT, you can set the repMatch option to a large value to try to improve finding matches in repetitive regions and do not use one of the default 11.ooc repeat masking files.
This usually occurs on the newer genome assemblies, such as hg38, when you search a sequence that has an "alternate" or "fix" sequence. To improve the quality of the these assemblies, curators have added multiple versions of some important loci, e.g. the MHC regions. They also add fix sequences to resolve errors without changing the reference. See our patches blog post for more information.
When you blat or isPCR a sequence which matches a chromosome location that also has a fix or alt sequence, you will see a match on the reference chromosome (e.g. "chr1") and another match on the patch sequence (e.g. chr1_KN196472v1_fix). In most cases it is safe to ignore the patch hit, as a human genome will not contain both the reference and alternate sequence at the same time. For more information on the specific kinds of patch sequences see our FAQ entry on the topic.
Due to the high demand on our Blat servers, we restrict service for users who programmatically query the BLAT tool or do large batch queries. Program-driven use of BLAT is limited to a maximum of one hit every 15 seconds and no more than 5,000 hits per day. Please limit batch queries to 25 sequences or less.
For users with high-volume Blat demands, we recommend downloading the BLAT tool for local use. For more information, see Downloading BLAT source and documentation.
BLAT source and executables are freely available for academic, nonprofit and personal use. Commercial licensing information is available on the Kent Informatics website.
BLAT source may be downloaded from http://hgdownload.soe.ucsc.edu/admin/ (located at /kent/src/blat within the most recent jksrci*.zip source tree). For BLAT executables, go to http://hgdownload.soe.ucsc.edu/admin/exe/ and choose your machine type.
Documentation on BLAT program specifications is available here. Note that the command-line BLAT does not return matches to U nucleotides in the query sequence.
We almost always expect small differences between the hgBLAT/gfServer and the stand-alone, command-line Blat. The best matches can be found using pslReps and pslCDnaFilter utilities. The web-based Blat is tuned permissively with a minimum cut-off score of 20, which will display most of the alignments. We advise deciding which filtering parameters make the most sense for the experiment or analysis. Often these settings will be different and more stringent than those of the web-based Blat. With that in mind, use the following settings to approximate the search results of the web-based Blat:
Note: There are cases where the gfServer/gfClient approach provide a better approximation of web results than standalone Blat. See the example below for an overview of this process.
standalone Blat:
blat -stepSize=5 -repMatch=2253 -minScore=20 -minIdentity=0
database.2bit query.fa output.psl faToTwoBit:
gfServer (this is how the UCSC web-based BLAT servers are configured):
gfServer start blatMachine portX -stepSize=5 -log=untrans.log
database.2bitgfServer start blatMachine portY -trans -mask -log=trans.log
database.2bit For enabling DNA/DNA and DNA/RNA matches, only the host, port and twoBit files are needed. The same port is used for both untranslated Blat (gfClient) and PCR (webPcr). You'll need a separate Blat server on a separate port to enable translated Blat (protein searches or translated searches in protein-space).
gfClient:
Notes on repMatch:
For more information about how to replicate the score and percent identity matches displayed by our web-based Blat, please see this BLAT FAQ.
For more information on the parameters available for BLAT, gfServer, and gfClient, see the BLAT specifications.
Using any -ooc option in BLAT, such as -ooc=11.ooc, speeds up searches similar to repeat-masking sequence. The 11.ooc file contains sequences determined to be over-represented in the genome sequence. To improve search speed, these sequences are not used when seeding an alignment against the genome. For reasonably sized sequences, this will not create a problem and will significantly reduce processing time.
By not using the 11.ooc file, you will increase alignment time, but will also slightly increase sensitivity. This may be important if you are aligning shorter sequences or sequences of poor quality. For example, if a particular sequence consists primarily of sequences in the 11.ooc file, it will never be seeded correctly for an alignment if the -ooc flag is used.
In summary, if you are not finding certain sequences and can afford the extra processing time, you may want to run BLAT without the 11.ooc file if your particular situation warrants its use.
There is no option to command-line Blat that gives you the percent ID and the score. However, we have created scripts that include the calculations:
pslScore.plpslScore.c
and associated library functions pslScore and pslCalcMilliBad in
psl.cSee our FAQ on source code licensing and downloads for information on obtaining the source.
The code for the "I'm feeling lucky" Blat search orders the results based on the sort output option that you selected on the query page. It then returns the highest-scoring alignment of the first query sequence.
If you are sorting results by "query, start" or "chrom, start", generating the "I'm feeling lucky" result is straightforward: sort the output file by these columns, then select the top result.
To replicate any of the sort options involving score, you first must calculate the score for each result in your PSL output file, then sort the results by score or other combination (e.g. "query, score" and "chrom, score"). See the section on Replicating web-based Blat percent identity and score calculations for information on calculating the score.
Alternatively, you can try filtering your Blat PSL output using either the
pslReps or
pslCDnaFilter program available in the Genome Browser source code. For information on
obtaining the source code, see our FAQ on source code licensing and
downloads.
Here are some guidelines for configuring standalone Blat and gfServer/gfClient for these conditions:
The above changes will make BLAT more sensitive, but will also slow the speed and increase the memory usage. It may be necessary to process one chromosome at a time to reduce the memory requirements.
A note on filtering output: increasing the -minScore parameter value beyond one-half of
the query size has no further effect. Therefore, use either the pslReps or
pslCDnaFilter program available in the Genome Browser source code to filter for the size,
score, coverage, or quality desired. For information on obtaining the source code, see our
FAQ on source code licensing and downloads.
BLAT is designed to quickly find sequence similarity between query and target sequences. Generally, BLAT is used to find locations of sequence homology in a single target genome or determine the exon structure of an mRNA. BLAT also allows users to compare the query sequence against all of the default assemblies for organisms hosted on the UCSC Genome Browser. The Search ALL feature may be useful if you have an ambiguous query sequence and are trying to determine what organism it may belong to.
Selecting the "Search ALL"
checkbox above the Genome drop-down list allows you to search the genomes
of the default assemblies for all of our organisms. It also searches any attached hubs'
Blat servers, meaning you can search your user-generated assembly hubs.
The new dynamic BLAT servers allow one to perform BLAT searches on an unlimited number of genomes with a fixed amount of memory,
however it takes time to swap virtual pages from the storage device.
Currently dynamic BLAT servers are not supported for "Search ALL", and they are noted as skipped in the output.
The results page displays an ordered list
of all our organisms and their homology with your query sequence. The results are ordered
so that the organism with the best alignment score is at the top, indicating which region(s)
of that organism has the greatest homology with your query sequence.
The entire alignment, including mismatches and gaps, must score
20 or higher in order to appear in the Blat output. By clicking into a link in the Assembly list
you will be taken to a new page displaying various locations and scores of sequence homology in the assembly of interest.
In the Blat ALL results page, the "Hits" column does not represent alignments, instead it reports tile hits. Tile hits are 11 base kmer matches found in the target, which do not necessarily represent successful alignments. When one clicks the 'Assembly' link a full Blat alignment for that genome will occur and any alignment scores representing less than a 20 bp result will come back as no matches found.
When you submit a sequence to the Blat ALL utility, the sequence is compared to an index in the server. The index has been built from the target genome, with an 11bp default stepSize. These 11-mers "tile" the sequence as such:
TGGACAACATG
GCAAGAATCAG
TCTCTACAGAA
After the index is built, the first step of alignment is to read the query (search) sequence, extract all the 11-mers, and look those up in the genome 11-mer index currently in memory. Matches found there represent the initial "hits" you see in the Blat ALL results page. The next step is to look for hits that overlap or fall within a certain distance of each other, and attempt to align the sequences between the hit locations in target and query.
For example, if two 11-base tile hits align perfectly, it would result in a score of 22. This is above the minimum required score of 20 (see Blat ALL genomes), and would be reported as an alignment. However, there are penalties for gaps and mismatches, as well as potential overlap (see stepsize in BLAT specifications), all of which could bring the score below 20. In that case, Blat ALL would report 2 "hits", but clicking into the assembly would report no matches. This most often occurs when there are only a few (1-3) hits reported by Blat ALL.
Often times using the gfServer/gfClient provides a better approximation or even replicate of the web-based Blat results, which otherwise cannot be found using standalone Blat. This approach mimics the Blat server used by the Genome Browser web-based Blat. The following example will show how to set up an hg19 gfServer, then make a query. First, download the appropriate utility for the operating system and give it executable permissions:
#For linux rsync -a rsync://hgdownload.soe.ucsc.edu/genome/admin/exe/linux.x86_64/blat/ ./ #For MacOS rsync -a rsync://hgdownload.soe.ucsc.edu/genome/admin/exe/macOSX.x86_64/blat/ ./ chmod +x gfServer gfClient blat
Next, download the appropriate .2bit genome (hg19 in this example), and run the gfServer utility with the web Blat parameters, designating the local machine and port 1234:
wget http://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.2bit ./gfServer start 127.0.0.1 1234 -stepSize=5 hg19.2bit
After a few moments, the gfServer will initialize and be ready to recieve queries. In order to approximate web Blat, we will use the gfClient with the following parameters, designating our input and output files.
./gfClient -minScore=20 -minIdentity=0 127.0.0.1 1234 . input.fa out.psl
The output file out.psl should have results very similar to web-based Blat.
This is due to how we store internal coordinates in the Genome Browser. The default Blat Output type of hyperlink shows results in our internal coordinate data structure. These internal coordinates have a zero-based start and a one-based end. See the following FAQ entry for more information.
If the Output type is changed to psl on web Blat, the same zero-based half open coordinate results will be seen as the standalone Blat and gfServer/gfClient procedures.
Protein-translated BLAT (protein or translated RNA queries) uses the standard vertebrate genetic code. It will be slightly less sensitive on mitochondria and species using other genetic codes. More information on standard genetic codes can be found on the NCBI website. Additional details on mitochondria codon tables can be found on the Wikiwand website.
For programmatic access, BLAT supports URL queries which are returned as psl format in a JSON structure. The URL requires three variables: the sequence to blat, the type of query and the database, as follows
https://genome.ucsc.edu/cgi-bin/hgBlat?userSeq=[seq]&type=[type]&db=[database]&output=json
If the sequence is longer than 8kbp, the request must be sent via HTTP POST.
Query types include DNA, protein, translated RNA and translated DNA:
https://genome.ucsc.edu/cgi-bin/hgBlat?userSeq=GACCTCGGCGTGGCCTAGCG&type=DNA&db=hg38&output=json https://genome.ucsc.edu/cgi-bin/hgBlat?userSeq=IGCLPAHLLGDMWGRFWTNLYSLTVPFGQKPNIDVTDAMVDQAWDAQRIFKEAEKFFVSVGLPNM&type=protein&db=hg38&output=json https://genome.ucsc.edu/cgi-bin/hgBlat?userSeq=UUUCCCUUCCCCACUGUAGUGGGAGAGAAGGGAGUGGCCAUACCAUAUUUUUCUCGUGGGCCGUUGUAGUCAUAAGGCCUUCCUUUGCGGAAAAUUUUCAGGGUGGGAUA&type=translated%20RNA&db=hg38&output=json https://genome.ucsc.edu/cgi-bin/hgBlat?userSeq=TTTCCCTTCCCCACTGTAGTGGGAGAGAAGGGAGTGGCCATACCATATTTTTCTCGTGGGCCGTTGTAGTCATAAGGCCTTCCTTTGCGGAAAATTTTCAGGGTGGGATA&type=translated%20DNA&db=hg38&output=json