Intro

We consider various mapping/relationship between gene, SNP, pathway etc and other biological annotations in order to get a biological insight, whatever it may mean in your particular experiment.

Genome Browsers and alike

Easiest way to annotate a small list of genes, SNPs, microarray probes etc is to query various web based interfaces often provided within genome browsers. Also useful when you need quick sanity check of your annotation scripts.

General interfaces

• UCSC genome browser
  • UCSC table browser
  • ENCODE at UCSC
• Ensembl genome browser
  • ENCODE at Ensembl
• NCBI
  • GTEx eQTL browser
  • NCBI Gene P53 example
  • NCBI SNP rs4783244 example
• Biomart
• Gencode human transcriptome annotation; here for the sake of completeness, part of the ENCODE project. Available as a track in the UCSC browser

SNP

• Haploreg2
• Gemini standalone tool
  • Browser interface
  • How to use plink files with gemini

Gene Expression

• BioGPS

Pathway Analysis

• DAVID

Biomart, UCSC, Ensembl and NCBI integrate broad spectrum of data within a unified interface, on the other hand there are plenty of annotation tools addressing particular needs e.g. eQTL browsers, miRNA and lincRNA databases not(yet) mentioned here.

Biomart

from www.biomart.org

BioMart is a freely available, open source, federated database system that provides unified access to disparate, geographically distributed data sources. It is designed to be data agnostic and platform independent, such that existing databases can easily be incorporated into the BioMart framework.

Ensembl is one of the databases available through Biomart. There are numerous APIs to access Ensembl programmatically. We're going to use R/Bioconductor interface which unlike other APIs allows access entire Biomart.

Ensembl/Biomart APIs:

• web interface Mart view
• R/Bioconductor biomaRt
• python: Pycogent does much more than just querying Ensembl, please look through the documentation
• ruby: ruby-ensembl-api
• perl: Perl API
• language agnostic beta REST API

It is instructive to retrieve some data using the highest level web-interface and the lowest MySQL level to see that after all there is no magic. The rest of APIs provide nice in-between level in your preferred language which hides all particularities of db schema but allows you to deal with the annotation programmatically.

Web interface a.k.a. build a query by selecting drop-down menus.

Hopefully this is self explanatory:

1. Go to Mart view
2. CHOOSE DATABASE: Ensembl Variation 73
3. CHOOSE DATASET: Homo Sapience Short Variation
4. Filters: GENERAL VARIATION FILTERS: Filter by Variation Name:
   • rs7775397
   • rs4783244
   • rs6450176
5. Attributes: SEQUENCE VARIATION: Variation Name, Minor allele (ALL), 1000 genomes global MAF (ALL)
6. Results

By the same token one can build a query via Biomart: result

Direct MySQL query

Query MySQL backend (not recommended, plus no sequence retrieval; schema available here)

mysql --host=ensembldb.ensembl.org -P 3306  --user=anonymous \
-A -e "SELECT v.name, v.minor_allele, v.minor_allele_freq \
FROM variation v WHERE v.name IN ('rs7775397', 'rs4783244', 'rs6450176');" \
homo_sapiens_variation_73_37

name    minor_allele    minor_allele_freq
rs4783244   T   0.3393
rs6450176   A   0.3356
rs7775397   G   0.033

Note

Before we go into discovering lots of goodies in the Bioconductor project. Bioconductor conference and tutorial materials are available here. Each package in the project has a vignette which is worth checking. Bioconductor also have workflows section describing particular usage of the project e.g. Using bioconductor for annotation.

biomaRt

library(biomaRt)

Available marts

Contrast ensembl versions available through web/MySQL interfaces and biomaRt.

kable(head(listMarts()))

Available datasets

Select snp mart and see which datasets are available for the mart.

mart <- useMart("snp")
kable(head(listDatasets(mart)))

Select dataset with human snps hsapiens_snp.

snpmart <- useDataset("hsapiens_snp", mart = mart)

Shortcut in case you know which mart and dataset you are after:

snpmart <- useMart("snp", dataset = "hsapiens_snp")

Attributes and Filters

A dataset queried on a set of fields: Attributes i.e. the desired output. A query narrowed based on a set of fields: Filters.

Here are Attributes for the hsapiens_snp dataset of the snp mart.

kable(head(listAttributes(snpmart)))

Here are the Filters

kable(head(listFilters(snpmart)))

Example query: Suppose we have a list of rs SNP ids (Filter, we want only data for those rs ids) and would like to figure out where those SNPs located, their minor alleles and MAFs (Attributes, we need only those fields returned from biomart). The getBM is the main function to query Biomart. We have already seen first four arguments: attributes, filters, value (of the filters) and mart.

?getBM

top_ids <- c("rs7775397", "rs4783244", "rs6450176")
snp_pos <- getBM(attributes = c("refsnp_id", "chr_name", "chrom_start",
                                 "minor_allele", "minor_allele_freq"),
                    filters = c("snp_filter"),
                      value = top_ids,
                       mart = snpmart)
kable(snp_pos)

All together

library(biomaRt)
snpmart <- useMart("snp", dataset = "hsapiens_snp")
top_ids <- c("rs7775397", "rs4783244", "rs6450176")
snp_pos <- getBM(attributes = c("refsnp_id", "chr_name", "chrom_start",
                                 "minor_allele", "minor_allele_freq"),
                    filters = c("snp_filter"),
                      value = top_ids,
                       mart = snpmart)

Full tables for biomaRt

Marts

kable(listMarts())

Datasets available for the snp mart

kable(listDatasets(mart))

Attributes for the snp mart, the hsapiens_snp dataset

kable(listAttributes(snpmart))

Filters for the snp mart, the hsapiens_snp dataset

kable(listFilters(snpmart))