TopHat

A spliced read mapper for RNA-Seq

      

Site Map

News and updates

    New releases and related tools will be announced through the Bowtie mailing list.

Getting Help

Releases

Related Tools

  • Cufflinks: Isoform assembly and quantitation for RNA-Seq
  • Bowtie: Ultrafast short read alignment
  • TopHat-Fusion: An algorithm for Discovery of Novel Fusion Transcripts
  • CummeRbund: Visualization of RNA-Seq differential analysis

Publications

Contributors

Links

Manual


What is TopHat?


TopHat is a program that aligns RNA-Seq reads to a genome in order to identify exon-exon splice junctions. It is built on the ultrafast short read mapping program Bowtie. TopHat runs on Linux and OS X.


What types of reads can I use TopHat with?


TopHat was designed to work with reads produced by the Illumina Genome Analyzer, although users have been successful in using TopHat with reads from other technologies. In TopHat 1.1.0, we began supporting Applied Biosystems' Colorspace format. The software is optimized for reads 75bp or longer.


How does TopHat find junctions?


TopHat can find splice junctions without a reference annotation. By first mapping RNA-Seq reads to the genome, TopHat identifies potential exons, since many RNA-Seq reads will contiguously align to the genome. Using this initial mapping information, TopHat builds a database of possible splice junctions and then maps the reads against these junctions to confirm them.

Short read sequencing machines can currently produce reads 100bp or longer but many exons are shorter than this so they would be missed in the initial mapping. TopHat solves this problem mainly by splitting all input reads into smaller segments which are then mapped independently. The segment alignments are put back together in a final step of the program to produce the end-to-end read alignments.

TopHat generates its database of possible splice junctions from two sources of evidence. The first and strongest source of evidence for a splice junction is when two segments from the same read (for reads of at least 45bp) are mapped at a certain distance on the same genomic sequence or when an internal segment fails to map - again suggesting that such reads are spanning multiple exons. With this approach, "GT-AG", "GC-AG" and "AT-AC" introns will be found ab initio. The second source is pairings of "coverage islands", which are distinct regions of piled up reads in the initial mapping. Neighboring islands are often spliced together in the transcriptome, so TopHat looks for ways to join these with an intron. We only suggest users use this second option (--coverage-search)  for short reads (< 45bp) and with a small number of reads (<= 10 million).  This latter option will only report alignments across "GT-AG" introns


Prerequisites


To use TopHat, you will need the following programs in your PATH:

  • bowtie2 and bowtie2-align (or bowtie)
  • bowtie2-inspect (or bowtie-inspect)
  • bowtie2-build (or bowtie-build)
  • samtools

Because TopHat outputs and handles alignments in BAM format, you will need to download and install the SAM tools. You may want to take a look at the Getting started guide for more detailed installation instructions, including installation of SAM tools and Boost.

You will also need Python version 2.6 or higher.


Obtaining and installing TopHat


You can download the latest source release and precompiled binaries for Linux and Mac OSX here. See the Getting started guide for detailed instructions about installing TopHat from the binary package or building TopHat and its dependencies from source.

To install TopHat from source package, unpack the tarball and change directory to the package directory as follows:

tar zxvf tophat-2.0.0.tar.gz
cd tophat-2.0.0/

Configure the package, specifying the install path and the library dependencies as needed (see the  Getting started guide for details):

./configure --prefix=<install_prefix> --with-boost=<boost_install_prefix> --with-bam=<samtools_install_prefix>

Finally, build and install TopHat:

make
make install

As detailed in the Getting started guide, if you want to install TopHat 2 without overwriting a previous version of TopHat already installed on your system you should specify a new, separate <install_prefix> for the ./configure command above, and after the 'make install' step just copy the tophat2 script from <install_prefix>/bin to a directory that is in your shell's PATH, so you can invoke this new version of TopHat with the command 'tophat2'.

Below you will find a detailed list of command-line options you can use to control TopHat. Beginning users should take a look at the Getting started guide for a tutorial on installing and running TopHat and its prerequisites.

Please Note TopHat has a number of parameters and options, and their default values are tuned for processing mammalian RNA-Seq reads. If you would like to use TopHat for another class of organism, we recommend setting some of the parameters with more strict, conservative values than their defaults. Usually, setting the maximum intron size to 4 or 5 Kb is sufficient to discover most junctions while keeping the number of false positives low.

Using TopHat


Usage: tophat [options]* <genome_index_base> <reads1_1[,...,readsN_1]> [reads1_2,...readsN_2]

When running TopHat with paired reads it is critical that the *_1 files an the *_2 files appear in separate comma-delimited lists, and that the order of the files in the two lists is the same.
TopHat allows the use of additional unpaired reads to be provided after the paired reads. These unpaired reads can be either given at the end of the paired read files on one side (as reads that can no longer be paired with reads from the other side), or they can be given in separate file(s) which are appended (comma delimited) to the list of paired input files on either side e.g.:

tophat [options]* <genome_index_base> PE_reads_1.fq.gz,SE_reads.fa PE_reads_2.fq.gz
or
tophat [options]* <genome_index_base> PE_reads_1.fq.gz PE_reads_2.fq.gz,SE_reads.fa
Starting with version 2.0.10 TopHat accepts mixed input file formats (FASTA/FASTQ).

NOTE: TopHat can align reads that are up to 1024 bp long, and it handles paired-end reads and unpaired reads at once, but we do not recommend mixing different types of reads in the same TopHat run. For example, mixing 100bp single end reads and 2x27bp paired reads in the same TopHat run may give sub-optimal results. If you'd like to combine results from data sets with different types of RNA-Seq reads, you can follow a protocol like this:

  • run TopHat on the first set of reads, with the appropriate parameters for this data set
  • use bed_to_juncs to convert the junctions.bed file obtained in this first run to a junction file usable by Tophat's -j option
  • run Tophat on the 2nd set of reads using the -j option to supply the junctions file produced by bed_to_juncs in the previous step

The following is a detailed description of the options used to control the TopHat script.

Arguments:
<genome_index_base> The basename of the genome index to be searched. The basename is the name of any of the index files up to but not including the first period. Bowtie first looks in the current directory for the index files, then looks in the indexes subdirectory under the directory where the currently-running bowtie executable is located, then looks in the directory specified in the BOWTIE_INDEXES  (or BOWTIE2_INDEXES) environment variable.
Please note that it is highly recommended that a FASTA file with the sequence(s) the genome being indexed be present in the same directory with the Bowtie index files and having the name <genome_index_base>.fa. If not present, TopHat will automatically rebuild this FASTA file from the Bowtie index files.
<reads1_1[,...,readsN_1]> A comma-separated list of files containing reads in FASTQ or FASTA format. When running TopHat with paired-end reads, this should be the *_1 ("left") set of files.
<[reads1_2,...readsN_2]> A comma-separated list of files containing reads in FASTA or FASTA format. Only used when running TopHat with paired end reads, and contains the *_2 ("right") set of files. The *_2 files MUST appear in the same order as the *_1 files.
Options:
-h/--help Prints the help message and exits
-v/--version Prints the TopHat version number and exits
-N/--read-mismatches Final read alignments having more than these many mismatches are discarded. The default is 2.
--read-gap-length Final read alignments having more than these many total length of gaps are discarded. The default is 2.
--read-edit-dist Final read alignments having more than these many edit distance are discarded. The default is 2.
--read-realign-edit-dist Some of the reads spanning multiple exons may be mapped incorrectly as a contiguous alignment to the genome even though the correct alignment should be a spliced one - this can happen in the presence of processed pseudogenes that are rarely (if at all) transcribed or expressed. This option can direct TopHat to re-align reads for which the edit distance of an alignment obtained in a previous mapping step is above or equal to this option value. If you set this option to 0, TopHat will map every read in all the mapping steps (transcriptome if you provided gene annotations, genome, and finally splice variants detected by TopHat), reporting the best possible alignment found in any of these mapping steps. This may greatly increase the mapping accuracy at the expense of an increase in running time. The default value for this option is set such that TopHat will not try to realign reads already mapped in earlier steps.
--bowtie1 Uses Bowtie1 instead of Bowtie2. If you use colorspace reads, you need to use this option as Bowtie2 does not support colorspace reads.
-o/--output-dir <string> Sets the name of the directory in which TopHat will write all of its output. The default is "./tophat_out".
-r/--mate-inner-dist <int> This is the expected (mean) inner distance between mate pairs. For, example, for paired end runs with fragments selected at 300bp, where each end is 50bp, you should set -r to be 200. The default is 50bp.
--mate-std-dev <int> The standard deviation for the distribution on inner distances between mate pairs. The default is 20bp.
-a/--min-anchor-length <int> The "anchor length". TopHat will report junctions spanned by reads with at least this many bases on each side of the junction. Note that individual spliced alignments may span a junction with fewer than this many bases on one side. However, every junction involved in spliced alignments is supported by at least one read with this many bases on each side. This must be at least 3 and the default is 8.
-m/--splice-mismatches <int> The maximum number of mismatches that may appear in the "anchor" region of a spliced alignment. The default is 0.
-i/--min-intron-length <int> The minimum intron length. TopHat will ignore donor/acceptor pairs closer than this many bases apart. The default is 70.
-I/--max-intron-length <int> The maximum intron length. When searching for junctions ab initio, TopHat will ignore donor/acceptor pairs farther than this many bases apart, except when such a pair is supported by a split segment alignment of a long read. The default is 500000.
--max-insertion-length <int> The maximum insertion length. The default is 3.
--max-deletion-length <int> The maximum deletion length. The default is 3.
--solexa-quals Use the Solexa scale for quality values in FASTQ files.
--solexa1.3-quals As of the Illumina GA pipeline version 1.3, quality scores are encoded in Phred-scaled base-64. Use this option for FASTQ files from pipeline 1.3 or later.
-Q/--quals Separate quality value files - colorspace read files (CSFASTA) come with separate qual files.
--integer-quals Quality values are space-delimited integer values, this becomes default when you specify -C/--color.
-C/--color Colorspace reads, note that it uses a colorspace bowtie index and requires Bowtie 0.12.6 or higher.
Common usage: tophat --color --quals [other options]* <colorspace_index_base> <reads1_1[,...,readsN_1]> [reads1_2,...readsN_2] <quals1_1[,...,qualsN_1]> [quals1_2,...qualsN_2]
-p/--num-threads <int> Use this many threads to align reads. The default is 1.
-g/--max-multihits <int> Instructs TopHat to allow up to this many alignments to the reference for a given read, and choose the alignments based on their alignment scores if there are more than this number. The default is 20 for read mapping. Unless you use --report-secondary-alignments, TopHat will report the alignments with the best alignment score. If there are more alignments with the same score than this number, TopHat will randomly report only this many alignments. In case of using --report-secondary-alignments, TopHat will try to report alignments up to this option value, and TopHat may randomly output some of the alignments with the same score to meet this number.
--report-secondary-alignments By default TopHat reports best or primary alignments based on alignment scores (AS). Use this option if you want to output additional or secondary alignments  (up to 20 alignments will be reported this way, this limit can be changed by using the -g/--max-multihits option above).
--no-discordant For paired reads, report only concordant mappings.
--no-mixed For paired reads, only report read alignments if both reads in a pair can be mapped (by default, if TopHat cannot find a concordant or discordant alignment for both reads in a pair, it will find and report alignments for each read separately; this option disables that behavior).
--no-coverage-search Disables the coverage based search for junctions.
--coverage-search Enables the coverage based search for junctions. Use when coverage search is disabled by default (such as for reads 75bp or longer), for maximum sensitivity.
--microexon-search With this option, the pipeline will attempt to find alignments incident to micro-exons. Works only for reads 50bp or longer.
--library-type The default is unstranded (fr-unstranded). If either fr-firststrand or fr-secondstrand is specified, every read alignment will have an XS attribute tag as explained below. Consider supplying library type options below to select the correct RNA-seq protocol.
Library TypeExamplesDescription
fr-unstrandedStandard IlluminaReads from the left-most end of the fragment (in transcript coordinates) map to the transcript strand, and the right-most end maps to the opposite strand.
fr-firststranddUTP, NSR, NNSRSame as above except we enforce the rule that the right-most end of the fragment (in transcript coordinates) is the first sequenced (or only sequenced for single-end reads). Equivalently, it is assumed that only the strand generated during first strand synthesis is sequenced.
fr-secondstrandLigation, Standard SOLiDSame as above except we enforce the rule that the left-most end of the fragment (in transcript coordinates) is the first sequenced (or only sequenced for single-end reads). Equivalently, it is assumed that only the strand generated during second strand synthesis is sequenced.
Advanced Options:
--bowtie-n TopHat uses "-v" in Bowtie for initial read mapping (the default), but with this option, "-n" is used instead. Read segments are always mapped using "-v" option.
--segment-mismatches Read segments are mapped independently, allowing up to this many mismatches in each segment alignment. The default is 2.
--segment-length Each read is cut up into segments, each at least this long. These segments are mapped independently. The default is 25.
--min-segment-intron The minimum intron length that may be found during split-segment search. The default is 50.
--max-segment-intron The maximum intron length that may be found during split-segment search. The default is 500000.
--min-coverage-intron The minimum intron length that may be found during coverage search. The default is 50.
--max-coverage-intron The maximum intron length that may be found during coverage search. The default is 20000.
--keep-tmp Causes TopHat to preserve its intermediate files produced during the run (mostly useful for debugging). The default is to delete these temporary files.
--keep-fasta-order In order to sort alignments in the same order in the genome fasta file, the option can be used. But this option will make the output SAM/BAM file incompatible with those from the previous versions of TopHat (1.4.1 or lower).
--no-sort-bam Output BAM is not coordinate-sorted.
--no-convert-bam Do not convert to bam format. Output is <output_dir>/accepted_hit.sam. Implies --no-sort-bam.
-R/--resume <string> In case a TopHat run was terminated prematurely (process failure due to external factors, e.g. running out of memory because of other processes running on the same machine, or the disk getting full), users can attempt to resume the interrupted TopHat run by just providing this option with the output directory for that run. TopHat sets several checkpoints after every lengthy operations in the pipeline and when this option is provided, it will attempt to resume the pipeline from the last successful checkpoint. This special usage of TopHat only requires this option, e.g. the command line could simply be:
    tophat -R tophat_out (or your TopHat output directory if you used the -o/--output-dir option)
Note that none of the original options used for the original TopHat run should be provided, TopHat will find all the original options (and the checkpoint info) in the logs/run.log file found in the specified directory.
-z/--zpacker Manually specify the program used for compression of temporary files; default is gzip; use -z0 to disable compression altogether. Any program that is option-compatible with gzip can be used (e.g. bzip2, pigz, pbzip2).

Bowtie 2 specific options:

Bowtie 2 provides many options so that users can have more flexibility as to how reads are mapped. TopHat 2 allows users to pass many of these options to Bowtie 2 by preceding the Bowtie 2 option name with the --b2- prefix.  Please refer to the Bowtie2 website for detailed information.

Preset options in --end-to-end mode  (local alignment is not used in TopHat2):
Tophat 2 option:
Corresponding Bowtie 2 option:
--b2-very-fast --very-fast
--b2-fast --fast
--b2-sensitive --sensitive
--b2-very-sensitive --very-sensitive
Alignment options:
--b2-N The default is 0.
--b2-L The default is 20.
--b2-i The default is S,1,1.25.
--b2-n-ceil The default is L,0,0.15.
--b2-gbar The default is 4.
Scoring options:
--b2-mp The default is 6,2.
--b2-np The default is 1.
--b2-rdg The default is 5,3.
--b2-rfg The default is 5,3.
--b2-score-min The default is L,-0.6,-0.6.
Effort options:
--b2-D The default is 15.
--b2-R The default is 2.
Fusion mapping options:

Reads can be aligned to potential fusion transcripts if the --fusion-search option is specified. The fusion alignments are reported in SAM format using custom fields XF and XP (see the output format) and some additional information about fusions will be reported (see fusions.out). Once mapping is done, you can run tophat-fusion-post to filter out fusion transcripts (see the TopHat-Fusion website for more details).

--fusion-search Turn on fusion mapping
--fusion-anchor-length A "supporting" read must map to both sides of a fusion by at least these many bases. The default is 20.
--fusion-min-dist For intra-chromosomal fusions, TopHat-Fusion tries to find fusions separated by at least this distance. The default is 10000000.
--fusion-read-mismatches Reads support fusions if they map across fusion with at most these many mismatches. The default is 2.
--fusion-multireads Reads that map to more than these many places will be ignored. It may be possible that a fusion is supported by reads (or pairs) that map to multiple places. The default is 2.
--fusion-multipairs Pairs that map to more than these many places will be ignored. The default is 2.
--fusion-ignore-chromosomes Ignore some chromosomes such as chrM when detecting fusion break points. Please check the correct names for chromosomes, that is, mitochondrial DNA is represented as chrM or M depending on the annotation you use.
Supplying your own transcript annotation data:

The options below allow you validate your own list of known transcripts or junctions with your RNA-Seq data. Note that the chromosome names in the files provided with the options below must match the names in the Bowtie index. These names are case-senstitive.



-j/--raw-juncs <.juncs file>

Supply TopHat with a list of raw junctions. Junctions are specified one per line, in a tab-delimited format. Records look like:


<chrom> <left> <right> <+/->

left and right are zero-based coordinates, and specify the last character of the left sequenced to be spliced to the first character of the right sequence, inclusive. That is, the last and the first positions of the flanking exons. Users can convert junctions.bed (one of the TopHat outputs) to this format using bed_to_juncs < junctions.bed > new_list.juncs where bed_to_juncs can be found under the same folder as tophat

--no-novel-juncs Only look for reads across junctions indicated in the supplied GFF or junctions file. (ignored without -G/-j)
-G/--GTF <GTF/GFF3 file>

Supply TopHat with a set of gene model annotations and/or known transcripts, as a GTF 2.2 or GFF3 formatted file. If this option is provided, TopHat will first extract the transcript sequences and use Bowtie to align reads to this virtual transcriptome first. Only the reads that do not fully map to the transcriptome will then be mapped on the genome. The reads that did map on the transcriptome will be converted to genomic mappings (spliced as needed) and merged with the novel mappings and junctions in the final tophat output.

Please note that the values in the first column of the provided GTF/GFF file (column which indicates the chromosome or contig on which the feature is located), must match the name of the reference sequence in the Bowtie index you are using with TopHat. You can get a list of the sequence names in a Bowtie index by typing:


bowtie-inspect --names your_index

So before using a known annotation file with this option please make sure that the 1st column in the annotation file uses the exact same chromosome/contig names (case sensitive) as shown by the bowtie-inspect command above.
--transcriptome-index <dir/prefix>

When providing TopHat with a known transcript file (-G/--GTF option above), a transcriptome sequence file is built and a Bowtie index has to be created for it in order to align the reads to the known transcripts. Creating this Bowtie index can be time consuming and in many cases the same transcriptome data is being used for aligning multiple samples with TopHat. A transcriptome index and the associated data files (the original GFF file) can be thus reused for multiple TopHat runs with this option, so these files are only created for the first run with a given set of transcripts. If multiple TopHat runs are planned with the same transcriptome data, TopHat should be first run with the -G/--GTF option together with the --transcriptome-index option pointing to a directory and a name prefix which will indicate where the transcriptome data files will be stored. Then subsequent TopHat runs using the same --transcriptome-index option value will directly use the transcriptome data created in the first run (no -G option needed after the first run).
Please note that starting with version 2.0.10 TopHat can be invoked with just the -G/--GTF and --transcriptome-index options but without providing any input reads (the <genome_index_base> argument is still required). This is a special usage directing TopHat to only build the transcriptome index data files for the given annotation and then exit.
Note: Only after the transcriptome files are built with one of the methods above, by a single TopHat process, it is safe to run multiple TopHat processes simultaneously making use of the same pre-built transcriptome index data.

For example, in order to just prepare the transcriptome index files for a specific annotation, an initial, single TopHat run could be invoked like this:
tophat -G known_genes.gtf \
    --transcriptome-index=transcriptome_data/known \
    hg19

In this example TopHat will create the transcriptome_data directory in the current directory (if it doesn't exist already) containing files known.gff, known.fa, known.fa.tlst, known.fa.ver and the known.* Bowtie index files. Then for subsequent TopHat runs with the same genome and known transcripts but different reads, TopHat will no longer spend time building the transcriptome index because it can use the previously built transcriptome index files, so the -G option is no longer needed (however using it again will not force TopHat to rebuild the transcriptome index files if they are already present and with the matching version)
tophat -o out_sample1 -p4 \
    --transcriptome-index=transcriptome_data/known \
    hg19 sample1_1.fq.z sample1_2.fq.z &

tophat -o out_sample2 -p4 \
    --transcriptome-index=transcriptome_data/known \
    hg19 sample2_1.fq.z sample2_2.fq.z &


The following options in this section are only used when the transcriptome search was activated with -G/--GTF and/or --transcriptome-index.
-T/--transcriptome-only Only align the reads to the transcriptome and report only those mappings as genomic mappings.
-x/--transcriptome-max-hits Maximum number of mappings allowed for a read, when aligned to the transcriptome (any reads found with more then this number of mappings will be discarded).
-M/--prefilter-multihits When mapping reads on the transcriptome, some repetitive or low complexity reads that would be discarded in the context of the genome may appear to align to the transcript sequences and thus may end up reported as mapped to those genes only. This option directs TopHat to first align the reads to the whole genome in order to determine and exclude such multi-mapped reads (according to the value of the -g/--max-multihits option).
Supplying your own insertions/deletions:

The options below allow you validate your own indels with your RNA-Seq data. Note that the chromosome names in the files provided with the options below must match the names in the Bowtie index. These names are case-senstitive.



--insertions/--deletions <.juncs file>

Supply TopHat with a list of insertions or deletions with respect to the reference. Indels are specified one per line, in a tab-delimited format, identical to that of junctions. Records look like:


Fordeletion,
<chrom> <left> <right>

left and right are zero-based coordinates, and specify the last character of the left sequenced to be spliced to the first character of the right sequence, inclusive.
For instance, "chr1 20564 20567", where two base pairs located at 20565 and 20566 are deleted in the sequenced genome.


For insertion,
<chrom> <left> <dummy> <inserted sequence>

left is zero-based coordinate and dummy can be set to the same value as left
. For instance, "chr1 17491 17491 CA", where two base pairs "CA" are inserted between 17490 and 17491 of the reference genome.

--no-novel-indels Only look for reads across indels in the supplied indel file, or disable indel detection when no file has been provided.

TopHat Output


The tophat script produces a number of files in the directory in which it was invoked. Most of these files are internal, intermediate files that are generated for use within the pipeline. The output files you will likely want to look at are:


  1. accepted_hits.bam. A list of read alignments in SAM format. SAM is a compact short read alignment format that is increasingly being adopted. The formal specification is here.
  2. junctions.bed. A UCSC BED track of junctions reported by TopHat. Each junction consists of two connected BED blocks, where each block is as long as the maximal overhang of any read spanning the junction. The score is the number of alignments spanning the junction.
  3. insertions.bed and deletions.bed. UCSC BED tracks of insertions and deletions reported by TopHat.
    Insertions.bed - chromLeft refers to the last genomic base before the insertion.
    Deletions.bed - chromLeft refers to the first genomic base of the deletion.