Genome browsing, Genomic data mining and Genome data visualization
with Ensembl, Biomart and IGV
Alex Sánchez
August 2005
2
What is Ensembl
• Ensembl is a joint scientific project between the European Bioinformatics Institute and the Wellcome Trust Sanger Institute, which was launched in 1999 in response to the imminent completion of the Human Genome Project.
• More than one decade later, Ensembl's aim remains to provide a centralized resource for geneticists, molecular biologists and other researchers studying the genomes of our own species and other vertebrates and model organisms.
• Ensembl is one of several well known genome browsers for the retrieval of genomic information.
3
“Ensembl” is a genome browser
• Ensembl provides a genome browser that acts as a single point of access to annotated genomes for mainly vertebrate species.
• Information such as gene sequence, splice variants and further annotation can be retrieved at the genome, gene and protein level. This includes information on – protein domains, genetic variation, homology, syntenic regions
and regulatory elements.
• Coupled with analyses such as whole genome alignments and effects of sequence variation on protein, this powerful tool aims to describe a gene or genomic region in detail.
4
Basic Genome Annotation
• Genes – Genomic location – Gene model structures
• Exons • Introns • UTRs
– Transcript(s) • Pseudogenes • Non-coding RNA
– Protein(s) – Links to other sources of information
5
Advanced Genome Annotation
• Cytogenetic bands
• Polymorphic markers
– Sequence Tagged Sites (STS)
• Genetic variation
– Single Nucleotide Polymorphisms (SNPs)
– Deletion-Insertion Polymorphisms (DIPs)
– Short Tandem Repeats (STRs)
• Repetitive sequences
• Expressed Sequence Tags (ESTs)
• cDNAs or mRNAs from related species
• Regions of sequence homology
6
Use ensembl if you:
• have a gene of interest, and you would like to know if there are homologues in other species, or any sequence variations in the gene;
• would like to know what the sequence is for your gene of interest, and what the sequences of the splice variants (transcripts) are;
• want to explore the region around a gene of interest, and find neighbouring genes;
• want to find sequences that may be involved in gene regulation (open chromatin signatures, transcription factor binding sites, etc.);
• are interested in how conserved a gene or region is across species;
• want to know a selection of sequence variants that have been associated with a disease, for example, diabetes;
• have questions about a gene, variant, or chromosomal region;
7
Don’t use it if:
• you want to submit sequence files (see the course on ENA);
• you are looking for metabolic pathways (learn more about Reactome);
• your species of interest is not a chordate (see a sister project, Ensembl Genomes);
The Ensembl web site
Ensembl … …
takes genomic sequence assemblies
human build 35, mouse, rat, mosquito…
adds annotation and links
automated process
presents all the data on a web site
9
How to search Ensembl
• Search www.ensembl.org using:
– a gene name (for example, BRCA2);
– a UniProtaccession number (for example, P51587);
– a disease name (for example, coronary heart disease);
– a variation (for example, rs1223);
– a location - a genomic region (for example, rat X:100000..200000);
– a PDBe ID or a Gene Ontology (GO) term
• Most search results will take you to the appropriate Ensembl view through a results page.
• If you search using a location you will be directed straight to the location tab (this tab provides a view of a region of a genome).
10
Example 1 Searching the BCRA2 gene
• Open the at www.ensembl.org
• Choose your species of interest (Human) using the pull-down menu to the left of the search box.
• Type in your search term of interest into the search box. In our example we are using the gene name 'BRCA2'.
– You could also use a UniProtKB accession number, for example 'P51587'.
• Click 'Go' to obtain the search results
• You should see the BRCA2 gene at the top of the list.
11
Searching the BCRA2 gene: Results
12
Searching the BCRA2 gene: Summary
13
Searching the BCRA2 gene: View
14
Searching sequences: BLAT/BLAST
• f you have a sequence, but you are not sure what the gene name or ID in Ensembl is, you can align it to the genome with BLAST or BLAT
• BLAT (The BLAST-Like Alignment Tool) is fast, but it demands more exact matches. BLAST will allow lower-scoring hits, and allows more gaps in alignments. You'll get more hits with BLAST (but it may be slower)
Searching Sequences Example: The MTAP4 gene
• CTCCGCACTGCTCACTCCCGCGCAGTGAGGTTGGCACAGCCACCGCTCTG TGGCTCGCTTGGTTCCCTTAGTCCCGAGCGCTCGCCCACTGCAGATTCCTT TCCCGTGCAGACATGGCCT
• Click on the BLAST/BLAT link at the top of the page (circled in red in figure).
• Paste your sequence into the box.
• Check the options are correct. For example, we have selected Homo sapiens as the species to search against and the BLAT search tool because we're looking for an identical match.
• Click 'Run'
Searching sequences: Results
• Alignment Locations vs. Karyotype. The alignment [A] shows all hits on the genome. The best hit is boxed. In this case, BLAT shows one hit.
• Alignment Locations vs. Query. The alignment [B] shows hits, or High Scoring Pairs (HSPs), as a red bar along the query sequence (the black and white bar below).
• Alignment Summary. The summary [C] shows a table of hits, with customisable columns. Links are provided from the table. The link 'A' shows an alignment of the query and target sequence. 'G' shows the hit on the genome. 'C' brings you to the location tab, where you can see the BLAT hit in context of genes in that region.
Regions, maps and markers
MarkerView
SNPView
GeneSNPView
ContigView
CytoView
SyntenyView
MultiContigView
Genes & gene products
GeneView
TransView
ExonView
ProteinView
FamilyView
DomainView
GOView
DiseaseView
Ensembl exercises
Type the name of your favorite gene (i.e. BRCA2) and explore all the sections of ensembl for this gene. •Has this gene an ortholog in mouse? •How many different transcript do we know of this gene? •How many exons has the longest transcript? •Which functional annotations has this gene? (hint: check at GO annotations •Can you find SNPs in this gene?
Data retrieval
BioMart
Data sets on ftp site
MySQL queries of databases
Perl API access to databases
Export View
35
ExportView
Data Mining in Ensembl with Biomart
August 2005
www.biomart.org/biomart/martview
37
Simple Text-based Search Engine
38
‘Mouse Gene’ Gives Us Results
39
A More Complex Query is Not as Useful
40
BioMart- Data mining
• BioMart is a search engine that can find multiple terms and put them into a table format.
• Such as: human gene (IDs), chromosome and base pair position
• No programming required!
41
General or Specific Data-Tables
• All the genes for one species
• Or… only genes on one specific region of a chromosome
• Or… genes on one region of a chromosome associated with a disease
42
BioMart Data Sets
• Ensembl genes
• Vega genes
• SNPs
• Markers
• Phenotypes
• Gene expression information
• Gene ontology
• Homology predictions
• Protein annotation
Web Interface
With BioMart, quickly extract gene-associated information from the Ensembl databases.
44
Information Flow
• Choose the species of interest (Dataset)
• Decide what you would like to know about the genes (Attributes)
(sequences, IDs, description…)
• Decide on a smaller geneset using Filters.
(enter IDs, choose a region …)
Web Interface
Three main stages: Dataset, Attributes and Filters.
Choose the species of interest
Choose what information
to view.
Choose the gene set using what
we know.
46
The First Step: Choose the Dataset
Homo sapiens genes are the
default.
47
The Second Step: Attributes
Attributes are what we want to know about the genes.
Four output pages.
48
The SNP Attribute Page
Output variation information such as SNP reference ID and alleles.
49
Filters Allow Gene Selection
Choose the gene set by region, gene ID(s), protein/domain type.
50
Export Sequence or Tables
Genes and attributes are exported as sequence (Fasta format) or tables.
51
Query:
• For all mouse genes on chromosome 10 that are protein coding, I would like to know the IDs in both Ensembl and MGI.
• In the query:
Attributes: what we want to know.
Filters: what we know
52
Query:
• For all mouse genes on chromosome 10 that are protein coding, I would like to know the IDs in both Ensembl and MGI.
• In the query:
Attributes: what we want to know.
Filters: what we know
53
Query:
• For all mouse genes on chromosome 10 that are protein coding, I would like to know the IDs in both Ensembl and MGI.
• In the query:
Attributes: what we want to know.
Filters: what we know
54
A Brief Example
Change dataset to mouse
Mus musculus
55
A Brief Example
Dataset has changed.
56
Attributes (Output Options)
Click Attributes.
Attributes allow us to choose what we wish to know.
IDs are found in the ‘Features’ page.
Click on ‘GENE’.
57
Default options selected: Ensembl Gene ID and Transcript ID
Attributes (Output Options)
Ensembl Gene ID is selected
58
Scroll down to select MGI symbol. Also select the accession number.
Attributes (Output Options)
‘Markersymbol ID’ will give us the MGI ID
59
‘Results’ give us Gene IDs for all mouse genes in the Ensembl database.
The Results Table
60
Select a Smaller Gene Set
Select ‘Filters’
Expand the REGION panel
Instead of all mouse genes, select protein coding genes on chromosome 10.
61
Select Genes on Chromosome 10
Select chromosome
10
Instead of all mouse genes, select protein coding genes on chromosome 10.
62
Select Protein Coding Genes
Filters are set to chromosome 10 and protein-coding genes. Genes must meet BOTH
criteria to be in the result table.
Gene type: protein coding
63
Results (Preview)
This is a preview- if you are happy with the table, click ‘Go’.
For the full result table: Go
64
Full Result Table
Ensembl Gene ID Transcript
ID MGI
symbol MGI Accession
Number
65
Original Query:
• For all mouse genes on chromosome 10 that are protein coding, I would like to know the IDs in both Ensembl and MGI.
• In the query:
Attributes: columns in the Result Table
Filters: what we know
66
Other Export Options (Attributes)
• Sequences: UTRs, flanking sequences, cDNA and peptides, etc
• Gene IDs from Ensembl and external sources (MGI, Entrez, etc.)
• Microarray data
• Protein Functions/descriptions (Interpro, GO)
• Orthologous gene sets
• SNP/ Variation Data
Central Server
www.biomart.org
WormBase
HapMap
Population frequencies
Inter- population comparisons
Gene annotation
DictyBase
Uniprot, MSD
72
GRAMENE
Rice, Maize, Arabidopsis genomes…
73
Integrated Genome Viewer
74
IGV can use and display many file
formats http://www.broadinstitute.org/software/igv/FileFormats
75
IGV: file formats, e.g. BAM
(binary version of SAM, or Sequence Alignment Formatted files)
76
77
Ask your question, and then gather the
data, the tools and hardware you need
• Data and Databases: you will take
workshops, you will read papers, and you
will go on-line: SeqAnswers & maybe the
bioinformatics.ca Links Directory
• Tools: you will take workshops, you will
read papers, and you will go on-line:
SeqAnswers & maybe the
bioinformatics.ca Links Directory
• Hardware: you need to decide?
What can you do with IGV?
Visualization of different genomic data types:
aligned sequence reads
mutations
copy number
RNA interference screens
gene expression
methylation and genomic annotations
List of supported data formats: http://www.broadinstitute.org/software/igv/FileFormats
For this example:
*.bam for the alignment file
*.gtf for the genome annotation data
Using IGV to visualize sequence alignment
and genomic annotations
Here we have selected hg18
because it was used for the
alignment
Step1: Choose the genome in the list
(or import your own genome file)
Using IGV to visualize sequence alignment
and genomic annotations
Sample files source:
http://manuals.bioinformatics.ucr.edu/home/gui-ngs-analysis
and ftp://ftp.broad.mit.edu/pub/igv/INMEGEN2010/
Step 2: Import your alignment file
File->Load from File
You can also download file from a URL, a DAS or a server
Using IGV to visualize sequence alignment
and genomic annotations
Step 2: Import your sequence alignment file
If you download a *.bam file, it must be sorted and indexed, and
the index *.bai file must be in the same directory
You can visualize several alignment files at the same time for the
same species
Using IGV to visualize sequence alignment
and genomic annotations
Step 3: select the data to display
You can either:
select a chromosome
select the coordinates
search for a gene
Using IGV to visualize sequence alignment
and genomic annotations
Step 4: visualize the read alignments on the sequence
You will not see the alignment if the region your are looking at in too large
for IGV: Zoom in using the + sign (in red) or by double-clicking on the
display area
double-click here to zoom in and see the alignment
Using IGV to visualize sequence alignment
and genomic annotations
Track
names
Genomic annotations (default: RefSeq)
Cytoband Genomic coordinates
Data panel
Using IGV to visualize sequence alignment
and genomic annotations
White reads: low alignment score
Other colors: depend on the color alignment code selected
(ex: insert size, pair orientation, read strand)
Annotated exons Annotated introns
Coverage of reads on the sequence
Using IGV to visualize sequence alignment
and genomic annotations
Reference sequence (here hg18)
2 examples of variation compared
to the reference sequence
Lighter color bases: low quality bases
Using IGV to visualize sequence alignment
and genomic annotations
Step 5.1: download genomic annotations file from UCSC table browser
Several ways of downloading gene annotation files can be used, for
example directly from the source sequence databases
1) Go on http://genome.ucsc.edu and click on Tables
Using IGV to visualize sequence alignment
and genomic annotations
Select the genome (here hg18)
Select the gene annotations
(here Ensembl)
Choose your file name and click on the “get output” button
Select the file format (here GTF)
Using IGV to visualize sequence alignment
and genomic annotations
Select File->Load from file and choose the GTF file you have
downloaded
You have know access to RefSeq and Ensembl gene
annotations:
Step 5.2: load the genomic annotation file in IGV
The more data and annotations you load, the more memory you need You can
select a higher memory threshold if you need it when you launch IGV
Using IGV to visualize sequence alignment
and genomic annotations
On this example you can visualize deletion (10kb, from IGV publication*)
Robinson et al., (2011) Nature Biotechnology 29: 24–26
Using IGV to visualize sequence alignment
and genomic annotations
You can also visualize copy number variation data (from IGV publication*)
Robinson et al., (2011) Nature Biotechnology 29: 24–26
92
Following OpenHelix, UCSC, & SeqAnswers
• OpenHelix
– http://www.openhelix.com/
– Twitter: @openhelix
– Blog: http://blog.openhelix.com/
• UCSC
– http://genome.ucsc.edu/
– Twitter: @GenomeBrowser
– More tutorials: http://genome.ucsc.edu/training.html
• SEQanswers
– Forum for NGS technologies http://seqanswers.com/