Website & database

Hardware & infrastructure

Previous infrastructure

• Originally two servers (dev & live) here at Hinxton
• The team could only control the dev server
• Releases had to be conducted by the web development team via email
• Servers were on old hardware which our systems team wanted to decommision

Slow development + no fail over + aging hardware

= high risk of lengthy denial of service

New external web architecture

Recent releases

Home page

• Text on the page is clearer to read and more concise
• New functional word cloud which activates a search for common root symbols
• New search bar within our masthead

Original HGNC Search

• HGNC "Quick Search" was totally made in house
• Suited the users needs
• However ... problems with scalability
• Difficult to maintain and reuse for other purposes

• Well known search platform used by many companies and widely used across campus
• Highly scalable & very quick
• Easy to maintain & separate from our code base
• Provides faceting and highlighting out of the box
• Can use wildcards, phrases and logical operators
• Can limit the search to one field

Gene symbol reports

• Our most important pages
• Slightly different layout and added some improvements
• New functionality: help information and references

Live demo

REST service

• Introduce a new REST service to retrieve data from within our database
• Built upon our Solr search server
• Users can return data within an XML or JSON format for easy parsing
• Three main commands:
• info
• search
• fetch

Other improvements

• Updated list search and renamed it "Multi-symbol checker"
• A tool for our users to check a multiple gene symbols
• Updated the interface
• New code to make the search quicker for large lists
• Downloads & statistics page
• Added statistics for genes which reside in alternative loci only
• Karyotype image controls the statistics per chromosome
• Data sets are no longer created on-the-fly and are stored within the EBI FTP site

Web statistics

All statistics were collected by Google analytics between

May 1, 2014 - Oct 28, 2014

Number of users

• Mean = 47,000
• Median = 53,000

Where do the users come from?

User behaviour

Who are our biggest referers?

Where do our users go?

In summary...

• We are widely used
• Steady number of users of 53,000 users per month
• Important role of linking resources together as well as providing gene nomenclature
• We are as widely used within North America (33%) as we are in Europe (34%).
• Asia (23%) an emerging user base