In 2016, aside from all the BioThings improvements happening behind-the-scenes, MyGene.info saw a number of improvements including:
-2016.07.07: API updated to v3
-2016.08.18: Python Client Updated to v.3.0.0
-2016.08.25: ExAC functional constraint scores added
-2016.11.30: API status added to site

Even more exciting was the incredible progress made by other groups that use MyGene.info's service, especially CIViC and MyGene2.

Clinical Interpretations of Variants in Cancer (CIViC) is an open data resource which demonstrates the power/utility of open data and crowdsourcing for cancer variant annotations. In addition to having a user-friendly interface that encourages community cancer variant annotation, the brains behind this valuable bioinformatics resource also committed to open data by switching to public licensing.

MyGene2 is simultaneously a means for crowdsourcing information on gene variants and a valuable matchmaking tool for the rare disease community. The site's ingenious use of open data for both researchers and patients alike have made it a finalist for the Open Science Prize (and if you can vote for it to win here)

2016 saw a large increase in MyGene.info usage which happily means that other researchers are finding the service useful and hopefully making progress in their important work. According to google analytics, MyGene.info's user increased to ~51 Million requests this year--over double the 19.2 Million recorded last year.

Other than the incredible work (done by others) using MyGene.info, the team has been busy this year sharing about their efforts:
-2016.01.22: Chunlei presents update for Heart BD2K technical conference call
-2016.03.29: MyGene.info/MyVariant.info paper accepted to Genome Biology
-2016.04.05: Kevin presents at ISCB's NGS 2016 conference
-2016.05.10: Corresponding press release to Eureka Alert and Science Daily
-2016.05.25: Genome Biology blog post
-2016.05.30: Rarecast podcast about mygene.info
-2016.06.09: Elasticsearch blog post
-2016.07.10: Kevin's poster presentation at ISMB 2016
-2016.07.12: Chunlei's presentation at ISMB 2016
-2016.11.29: Chunlei's presentation at BD2K AHM 2016
-2016.11.28-30: Kevin's poster presentation at BD2K AHM 2016

Here's looking forward to an exciting 2017...we can't wait to see what you do!

ExAC also released a computed "functional gene constraint" metrics for each transcript containing variants from ExAC dataset. Since this is at transcript/gene level, we have now imported this data into MyGene.info, instead of MyVariant.info (which is designed as variant-specific). You can now access ExAC's latest release of functional gene constraint data from our v3 API of MyGene.info (under the field of exac). You can find out the definitions of all sub-fields under exac, or learn more about the functional gene constraint data from their FAQ and README files.

Ready to try pulling this new data? Try the following query examples:

• Querying with a gene id (NCBI gene id or Ensembl gene id)

  http://mygene.info/v3/gene/1017?fields=exac

  http://mygene.info/v3/gene/ENSG00000153230?fields=exac

• Querying with a transcript id:

  http://mygene.info/v3/query?q=exac.transcript:ENST00000266970&fields=exac

If you use our python client, you'll be pleased to know that the most recent update of our client also enables you to run similar queries.

Python example:

In [1]: import mygene

In [2]: mg = mygene.MyGeneInfo()

In [3]: mg.getgene(1017, fields='exac')  
In [4]: mg.getgene('ENSG00000153230', fields='exac')  
In [5]: mg.query('exac.transcript:ENST00000266970', fields='exac')  

Just want to mention again this new exac field is only available from our v3 MyGene.info API, not available from the still-live v2 API. Just another reason to switch to v3 API now :-).

Enjoy!

Our Python client mygene module is now updated to use our v3 API as the default. We increased its version from v2.3.0 to v3.0.0, just to match the underlying API version.

The mygene module itself has no incompatible changes (see detailed CHANGES.txt if you like). But you might want to check out the changes in our underlying v3 API from this detailed migration guide, especially if you are using refseq, accession and exon fields in your application.We encourage all of our users to upgrade to this new version. The upgrade is as easy as one line of command:

pip install mygene -U  

To verify you have the latest version installed:

In [1]: import mygene

In [2]: mygene.__version__  
Out[2]: '3.0.0'  

And you can see it's using our v3 API as the default now:

In [3]: mg = mygene.MyGeneInfo()  
In [4]: mg.url  
Out[4]: 'http://mygene.info/v3'  

As a side note, in case you still want to use our v2 URL for a while, you can still do that by setting mg.url='http://mygene.info/v2'. Our v2 API will still be live till most of our users are migrated to v3 API, but the annotation data from v2 API will not be updated any more.

As always, you can find more info about our MyGene.info Python client here:

• MyGene.py PyPI page.
• The detailed API documentation
• Feedbacks? help@mgene.info or @mygeneinfo.

Here's a brief list of changes, we'll discuss some of them in depth in next posts:

Refseq accession number with version

As some of you requested this feature (see We need your opinion), we now store accession number with version. You can search this information with and without the version, so the following requests will give the same results:

• http://mygene.info/v3/query?q=NM_001798.4&fields=refseq (including version .4)
• http://mygene.info/v3/query?q=NM_001798&fields=refseq (no version)
• http://mygene.info/v3/query?q=refseq.rna:NM_001798.4&fields=refseq (explicit refseq.rna key with version)
• http://mygene.info/v3/query?q=refseq.rna:NM_001798&fields=refseq (explicit refseq.rna key, no version)

...
  "refseq": {
    "genomic": [
      "NC_000012.12",
      "NC_018923.2",
      "NG_034014.1"
    ],
    "protein": [
      "NP_001277159.1",
      "NP_001789.2",
      "NP_439892.2",
      "XP_011536034.1"
    ],
    "rna": [
      "NM_001290230.1",
      "NM_001798.4",
      "NM_052827.3",
      "XM_011537732.1"
    ],
...

Note: v2 doesn't store version, see http://mygene.info/v2/query?q=refseq.rna:NM_001798&fields=refseq

RNA-protein mapping

"refseq", "accession" and "ensembl" now contains association between RNA and its protein product, within an added inner key "translation", as show in the following example for gene ID 1017.

Note: if a RNA or protein accession number isn't available in the association, then it's not added to this list

http://mygene.info/v3/gene/1017?fields=refseq

{
    "_id": "1017",
    "refseq": {
        ...
        "translation": [
          {   
            "protein": "XP_011536034.1",
            "rna": "XM_011537732.1"
          },  
          {   
            "protein": "NP_001789.2",
            "rna": "NM_001798.4"
          },  
          {   
            "protein": "NP_439892.2",
            "rna": "NM_052827.3"
          },  
          {   
            "protein": "NP_001277159.1",
            "rna": "NM_001290230.1"
          }   
        ]   
    }   
}

_Note: v2 does provide this information, see http://mygene.info/v2/gene/1017?fields=refseq

"exons" inner structure

Inner structure is now a list of dictionary. Each dictionary contains information about the exons with a "transcript" key containing the accession number. "position" inner key contains the different exons' positions.

http://mygene.info/v3/gene/1017?fields=exons

{                                                                                                                                                                                                                                                                                                                                                                           
    "_id": "1017",
    "_score": 21.731894,
    "exons": [
    {   
        "cdsend": 55971625,
        "cdsstart": 55967008,
        "chr": "12",
        "position": [
          [   
            55966768,
            55967124
          ],  
          [   
            55968048,
            55968169
          ],  
          [   
            55968777,
            55968948
          ],  
          [   
            55971043,
            55971247
          ],  
          [   
            55971520,
            55972789
          ]   
        ],  
        "strand": 1,
        "transcript": "NM_001290230",
        "txend": 55972789,
        "txstart": 55966768
    },  
    ...
}                      

Note: you can compare this structure with the actual v2, which uses a dictionary instead of a list of dictionary: http://mygene.info/v2/gene/1017?fields=exons

Better mapping between Ensembl and Entrez gene IDs

There are some annoying cases of one-to-many matches between Ensembl IDs and Entrez IDs, based on the mapping from Ensembl. For example, Ensembl gene ID ENSMUSG00000071350 associated to Entrez gene IDs 628705 and 239122. While these ambiguous mappings won't disappear completely, majority of them can be fixed by cross-checking the mappings from other sources. We worked hard to improve this mapping and remove discrepancy as much as we could. We'll post more about this soon.

• http://mygene.info/v3/query?q=ensembl.transcript:ENSMUST00000095775 returns only one result, for gene ID 239122, after disambiguation based on symbol Setdb2
• the same request http://mygene.info/v2/query?q=ensembl.transcript:ENSMUST00000095775 on v2 returns two results, for gene 239122 and 628705.

Querying "reporter" data source

Because some "reporter" IDs are integers (e.g. Affymetrix HuGene_1-1 array), just like Entrez gene IDs, "reporter" field now needs to be explicit in the query to avoid any confusion:

http://mygene.info/v3/query?q=reporter:2845421&fields=reporter

Change in dot.field notation default

The "dot.field" notation is when nested keys are returned using dot, like ["refseq.rna"], instead of nested structure, such as ["refseq"]["rna"]. This behavior can be triggered using dotfield=1 in conjunction with fields parameters. Default is now results are returned using nested structure, unless dotfield=1 is explicitly specified.

• http://mygene.info/v2/gene/1017?fields=refseq.rna will show dotfield notation:

{
  "_id": "1017",
  "refseq.rna": [
    "NM_001290230",
    "NM_001798",
    "NM_052827",
    "XM_011537732"
  ]
}

• http://mygene.info/v3/gene/1017?fields=refseq.rna will show nested structure

{
  "_id": "1017",
  "_score": 21.731894,
  "refseq": {
    "rna": [
      "NM_001290230.1",
      "NM_001798.4",
      "NM_052827.3",
      "XM_011537732.1"
    ]
  }
}

Note: this change is only for annotation endpoint /gene. Query endpoint /query already defaults to nested structure.

We focus on your needs so you're more than welcome to give feedback, comment any of these changes and request more. Again, stay tuned for more about this new version!

But the Genome Biology paper only covers a fraction of what we've done with MyGene.info and how we'd like to improve it. We have big plans for improving our service, such as:

1. Add more data resources. Although our service already provides simple programmatic access to almost 250 gene and protein annotation fields, there are plenty of other data resources we'd love to add! Do you like using MyGene.info, but are frustrated that a common data resource is not available? We'd like to change that with the addition of 20+ data resources which are on our shortlist.

2. Add more data types (not specifically for MyGene.info, but we're interested in being more useful by branching out.) If you use our service, you probably already know about our sister service, MyVariant.info, for genetic variant annotation data. We'd like to create a Software Development Kit (SDK) called Biothings.io, and apply what we've learned in creating and maintaining MyGene.info and MyVariant.info towards the creation of a similar resource for diseases and chemical compounds.

3. Help users keep even more up-to-date. We'd like to create a new application that will allow our users to get customizable and integrated notifications when something new is discovered or annotated--not just for genes, but other BioThings as well!

That's what we'd love to accomplish in the future.

Unfortunately, funding is uncertain and our grant is up for renewal. If you'd like to continue using MyGene.info, or would like to see similar sister projects for diseases, or chemicals in the future, please consider writing us a letter of support for our renewal application. The deadline is looming, so any help you can offer us would be very much appreciated.

Please see our post on the Su Lab blog for more details.

Regardless of our uncertain funding status, v3 of the MyGene.info API is currently under active development, and is expected to bring many new enhancements based on your requests. We expect to release v3 for beta testing very soon, so stay tuned!

If you plan to help, we need those letters by July 1st!

Recently, a user brought such an issue to our attention.

@mygeneinfo Would it be possible to include version numbers for RefSeq accessions (`refseq.*` values and `exons` keys)? Awesome service btw!

— Martijn Vermaat (@martijnvermaat) February 16, 2016

@martijnvermaat yep, heard this request from other users as well. create a ticket for us: https://t.co/TceQcCFIkz? @mygeneinfo

— Chunlei Wu (@chunleiwu) February 16, 2016

The suggestion/request was then formalized with the creation of an issue in our bitbucket repository.

@chunleiwu @mygeneinfo Sure, here it is. Thanks for considering! https://t.co/2Ef7zDUJf3

— Martijn Vermaat (@martijnvermaat) February 16, 2016

But before we spend the time and resources developing a new feature, we need to know two things! First, is it a feature other users would find valuable? Secondly, how would it best be implemented? To answer these questions, we need your opinion!

(Can't see the form? Fill it out here)

pip install mygene -U  

To verify you have the latest version installed:

In [1]: import mygene

In [2]: mygene.__version__  
Out[2]: '2.3.0'  

What's new and changed:

• new get_fields method to search for matching field names.

  Wonder which field(s) provides associated KEGG pathways for your favorite gene? Using mv.get_fields('kegg') will give you the answer. This is essentially the same you can do from the "available fields" table of our documentation, but from the Python client directly.

In [3]: mg = mygene.MyGeneInfo()

In [4]: mygene.get_fields("kegg")  
Out[4]:  
{u'pathway.kegg': {u'indexed': False, u'type': u'object'},
 u'pathway.kegg.id': {u'indexed': True, u'type': u'string'},
 u'pathway.kegg.name': {u'indexed': True, u'type': u'string'}}

# you can then pass it to fields parameter
In [5]: mg.getgene('1017', fields='pathway.kegg')  
# or
In [6]: mg.query('symbol:CDK2', fields='pathway.kegg')  

• Added a new fetch_all parameter for query method to retrieve large query.

  This feature enables users to stream all matching hits from a large query as a Python generator. Suppose you want to get back all human kinase genes, you can easily do it via this new fetch_all parameter:

# Normally, you can do the query like this:
In [7]: kinases = mg.query('name:kinase', species='human')  
# this will return top 10 hits of total 1073 hits
Out [7]: <output omitted here>

# Previously, to get all 1073 hits, you can use 
# "size" and "skip" parameters for paging:
# this is first 1000
In [8]: kinases = mg.query('name:kinase', species='human', size=1000)  
# this is the rest of 73
In [9]: kinases = mg.query('name:kinase',  species='human', size=1000, skip=1000)

# Although this works, but becomes harder for
# even larger queries. Using "fetch_all" parameter,
# you can handle this in an elegant way:
In [9]: kinases = mg.query('name:kinase', species='human', fetch_all=True)  
In [10]: kinases  
Out [10]" <generator object _fetch_all at 0x7fec027d2eb0>

# kinases is a Python generator, now you can
# loop through it to get all 1073 hits:
In [11]: for gene in kinases:  
   ....:     print gene['_id'], gene['symbol']

• getgene method now returns None if the input geneid does not match a known gene. Previously it raises an exception.

• Two methods for batch queries, getgenes and querymany, now accept anything iterable as the input, like a list, a tuple, or a generator. Previously, they only accept a list or a tuple.

• Finally, one major under-the-hook change is that we switched to requests Python module from httplib2 module for making underlying web service calls. While both are excellent modules, requests has gained popularity recently, so there's a good chance you'll already have it installed in your Python environment when you install mygene.

As always, you can find more info about our MyGene.info Python client here:

• MyGene.py PyPI page.
• The detailed API documentation
• Feedbacks? help@mgene.info or @mygeneinfo.

As an up-to-date gene annotation query service, MyGene.info API already handles over 2.5 million queries/month for researchers around the world. Building on the success of MyGene.info, we have now launched MyVariant.info, an equivalent service for up-to-date annotations on variants. Whether you want to retrieve annotations for observed variants (using genomic-based HGVS names), or query for known variants matching your criteria (e.g. those associated with genes, diseases), MyVariant.info allows you to do this and more. Use MyVariant.info to make queries across multiple annotation types.

Currently, MyVariant.info already includes annotations of over 280 million variants from twelve data sources, including popular dbNSFP, ClinVar, EVS, COSMIC, and dbSNP. And we are continuing expanding our coverage of variant annotations, e.g. ExAC data will soon be available from MyVariant.info.

Visit http://myvariant.info to learn more or try the live API here.

Funded! @mygeneinfo and @myvariantinfo projects just received @NIH_BD2K award for targeted software dev: http://t.co/kDPcKgUYPl

— Chunlei Wu (@chunleiwu) June 2, 2015

One of popular use cases for MyGene.info API is to make gene queries from the web applications directly via AJAX calls. Developers don't need to maintain their own gene annotation database to handle these queries. Gene annotation data served from MyGene.info API are public, usually you don't need to encrypt them when making an API call. However, if you have a SSL-enabled web application (i.e., URL starts with https:// instead of http://) and try to make AJAX call from this app to uncrypted MyGene.info API, you will hit a mixed content error. Moden browsers will just block the access to any uncrypted request.

With the support of SSL, now you can switch the URL of your MyGene.info query to use https instead. That way, all the traffic between MyGene.info and the client are encrypted and user browsers will be happy, without popping up an error.

Here are examples of both http and https requests:

• Regular http request without encryption:

GET http://mygene.info/v2/query?q=symbol:cdk2&species=human

HTTP/1.1 200 OK  
Content-Length: 183  
Content-Type: application/json; charset=UTF-8

{
    "hits": [
        {
            "_id": "1017", 
            "_score": 89.21377, 
            "entrezgene": 1017, 
            "name": "cyclin-dependent kinase 2", 
            "symbol": "CDK2", 
            "taxid": 9606
        }
    ], 
    "max_score": 89.21377, 
    "took": 6, 
    "total": 1
}

• https request with encryption:

GET https://mygene.info/v2/query?q=symbol:cdk2&species=human

HTTP/1.1 200 OK  
Content-Length: 183  
Content-Type: application/json; charset=UTF-8

{
    "hits": [
        {
            "_id": "1017", 
            "_score": 89.21377, 
            "entrezgene": 1017, 
            "name": "cyclin-dependent kinase 2", 
            "symbol": "CDK2", 
            "taxid": 9606
        }
    ], 
    "max_score": 89.21377, 
    "took": 6, 
    "total": 1
}

Have fun!

Just like what we did for human genes a while ago, supporting hg19 while upgrading to hg38, we now announce that mm9 is supported by MyGene.info for mouse genes, in addition to default mm10 we already support. Similar to our hg19 support, you can now get genomic locations and exon coordinations based on mm9 for mouse genes. And you can also query genes using mm9-based genomic intervals. Below are a few examples:

• To get the genomic location of a given gene based on mm9:

# http://mygene.info/v2/gene/12566?fields=genomic_pos_mm9
In [1]: mg.getgene(12566, fields='genomic_pos_mm9')  
Out[1]:  
{'_id': '12566',
 'genomic_pos_mm9': {
    'chr': '10',
    'start': 128134995,
    'strand': -1,
    'end': 128142107}
}

• To get the exon coordinates based on mm9:

# http://mygene.info/v2/gene/12566?fields=exons_mm9
In [2]: mg.getgene(12566, fields='exons_mm9')  
Out[2]:  
{'_id': '12566',
 'exons_mm9': {'NM_183417': {
   'cdsstart': 128136161,
   'txstart': 128134994,
   'cdsend': 128141883,
   'txend': 128142107,
   'chr': '10',
   'strand': -1,
   'exons': [[128134994, 128136266],
    [128136527, 128136731],
    [128137031, 128137175],
    [128138398, 128138500],
    [128139468, 128139639],
    [128140762, 128140883],
    [128140981, 128141059],
    [128141767, 128142107]]},
  'NM_016756': {
   'cdsstart': 128136161,
   'txstart': 128134994,
   'cdsend': 128141883,
   'txend': 128142107,
   'chr': '10',
   'strand': -1,
   'exons': [[128134994, 128136266],
    [128136527, 128136731],
    [128138398, 128138500],
    [128139468, 128139639],
    [128140762, 128140883],
    [128140981, 128141059],
    [128141767, 128142107]]}}
}

• To make genomic interval query based on mm9, you just need to add mm9. prefix:

# http://mygene.info/v2/query?q=mm9.chr12:57795963-57815592&species=mouse
In [3]: mg.query('mm9.chr12:57,795,963-57,815,592', species='mouse')  
Out[3]:  
{'took': 9,
 'hits': [{'_score': 8.126015,
   'name': 'paired box 9',
   '_id': '18511',
   'entrezgene': 18511,
   'symbol': 'Pax9',
   'taxid': 10090},
  {'_score': 8.086747,
   'name': 'solute carrier family 25 (mitochondrial oxodicarboxylate carrier), member 21',
   '_id': '217593',
   'entrezgene': 217593,
   'symbol': 'Slc25a21',
   'taxid': 10090}],
 'max_score': 8.126015,
 'total': 2}

As we mentioned above, mm10 is still the default genome assembly for mouse genes, so all of the following queries will return data based on mm10:

# http://mygene.info/v2/gene/12566?fields=genomic_pos
In [4]: mg.getgene(12566, fields='genomic_pos')

# http://mygene.info/v2/gene/12566?fields=exons
In [5]: mg.getgene(12566, fields='exons')

# http://mygene.info/v2/query?q=chr12:57795963-57815592&species=mouse
In [6]: mg.query('chr12:57,795,963-57,815,592', species='mouse')  

GET http://mygene.info/v2/species/<taxid>  

This API is simple. Here is a example:

request:

GET http://mygene.info/v2/species/9606  

response:

HTTP/1.1 200 OK  
Content-Length: 487  
Content-Type: application/json; charset=UTF-8

{
    "_id": 9606, 
    "authority": [
        "homo sapiens linnaeus, 1758"
    ], 
    "common_name": "man", 
    "genbank_common_name": "human", 
    "has_gene": true, 
    "lineage": [
        9606, 
        9605, 
        207598, 
        9604, 
        314295, 
        9526, 
        314293, 
        376913, 
        9443, 
        314146, 
        1437010, 
        9347, 
        32525, 
        40674, 
        32524, 
        32523, 
        1338369, 
        8287, 
        117571, 
        117570, 
        7776, 
        7742, 
        89593, 
        7711, 
        33511, 
        33213, 
        6072, 
        33208, 
        33154, 
        2759, 
        131567, 
        1
    ], 
    "parent_taxid": 9605, 
    "rank": "species", 
    "scientific_name": "homo sapiens", 
    "taxid": 9606, 
    "uniprot_name": "homo sapiens"
}

A few fields from the returned object are explained below, and the rest should be self-explanatory:

• has_gene: a flag to indicate if this taxonomy node is associated with any gene.
• lineage: the list of taxonomy nodes traversing from the current node to the root of the taxonomy tree.
• uniprot_name: organism name from uniprot

Two optional query parameters are available:

• include_children: if passed as “true” or “1”, an additional "children" field will be returned, containing all children nodes of the current taxonomy node.
• has_gene: if combined with "include_children=true" and passed as "true" or "1", returned "children" field will be filtered for children nodes associated with genes.

You can try the following examples by yourself:

GET http://mygene.info/v2/species/1239?include_children=true

GET http://mygene.info/v2/species/1239?include_children=true&has_gene=true  

As you can guess, the returned "children" fields with "has_gene=true" are exactly the taxid filter passed to MyGene.info when you query genes with "include_tax_tree" parameter turned on.

Please also note that the "children" field will be truncated if the list is greater than 10,000.

Credits:

This feature is made possible through the project led by Greg Stupp at recent 2nd NoB Hackathon.

In a typical MyGene.info query, you can query genes for one or multiple species by providing a "species" parameter:

http://mygene.info/v2/query?q=cdk2&species=human

http://mygene.info/v2/query?q=cdk2&species=9606,10090  

MyGene.info now allows you to query genes at the level beyond species, that is, you can now query for matching genes for any given genus, family, or even phylum from the taxonomy tree (effectively any node from the tree). For example, you can now query for "lytic enzyme" in any firmicutes (gram-positive bacteria, taxonomy id: 1239):

http://mygene.info/v2/query?q=lytic enzyme&species=1239&include_tax_tree=true  

or, in Python:

mg = mygene.MyGeneInfo()  
mg.query('lytic enzyme', species=1239, include_tax_tree=True)  

Note that include_tax_tree=true parameter toggles the query against all taxonomy ids under 1239 node in the taxonomy tree (including 1239 itself). As comparison, the query without this parameter:

http://mygene.info/v2/query?q=lytic enzyme&species=1239  

will return empty hits, as no genes are annotated at the level of firmicutes.

We expect this new feature will be particularly useful for the fields like evolutionary biology and microbiome. In fact, this was a requested feature from our users in those fields. So, please give a try and let us know your feedback.

The exact usage of this feature is summarized below:

• species parameter accepts one or multiple taxnomony ids (multiple ids separated by commas)

• passing include_tax_tree=true expands the query against any sub-nodes of passed taxids (from species parameter) in the taxonomy tree, including the taxids themselves.

• Since you can pass any taxonomy id from the taxonomy tree, we will cap the expanded taxonomy id list to 10,000, so that it won't overload our servers.

Credits:

This feature is made possible through the project led by Greg Stupp at recent 2nd NoB Hackathon.

• genomic_pos field is now based on hg38

  This field contains the genomic location of the given gene, the start/end positions are now based on hg38:

In [1]: mg.getgene('1017', fields='genomic_pos')  
Out[1]:  
{'_id': '1017',
 'genomic_pos': {'chr': '12',
  'end': 55972784,
  'start': 55966769,
  'strand': 1},
}

• exons field is now based on hg38

  This field contains the genomic locations of exons, as well as cdsstart/cdsend, txstart/txend data. All these values are now based on hg38:

In [2]: mg.getgene('1017', fields='exons')  
Out[2]:  
{'_id': '1017',
 'exons': {'NM_001290230': {'cdsend': 55971625,
   'cdsstart': 55967008,
   'chr': '12',
   'exons': [[55966768, 55967124],
    [55968048, 55968169],
    [55968777, 55968948],
    [55971043, 55971247],
    [55971520, 55972789]],
   'strand': 1,
   'txend': 55972789,
   'txstart': 55966768},
  'NM_001798': {'cdsend': 55971625,
   'cdsstart': 55967008,
   'chr': '12',
   'exons': [[55966768, 55967124],
    [55967856, 55967934],
    [55968048, 55968169],
    [55968777, 55968948],
    [55969474, 55969576],
    [55971043, 55971247],
    [55971520, 55972789]],
   'strand': 1,
   'txend': 55972789,
   'txstart': 55966768},
  'NM_052827': {'cdsend': 55971625,
   'cdsstart': 55967008,
   'chr': '12',
   'exons': [[55966768, 55967124],
    [55967856, 55967934],
    [55968048, 55968169],
    [55968777, 55968948],
    [55971043, 55971247],
    [55971520, 55972789]],
   'strand': 1,
   'txend': 55972789,
   'txstart': 55966768}}}

• Genome interval query is now based on hg38 by default:

In [3]: mg.query('chrX:151,073,054-151,383,976', species='human')  
Out[3]:  
{'hits': [{'_id': '100422930',
   '_score': 5.1352987,
   'entrezgene': 100422930,
   'name': 'microRNA 4330',
   'symbol': 'MIR4330',
   'taxid': 9606},
  {'_id': 'ENSG00000228717',
   '_score': 5.1352987,
   'symbol': 'AF013593.1',
   'taxid': 9606},
  {'_id': 'ENSG00000278724',
   '_score': 5.1352987,
   'name': 'Metazoan signal recognition particle RNA',
   'symbol': 'Metazoa_SRP',
   'taxid': 9606},
  {'_id': '9248',
   '_score': 5.1264653,
   'entrezgene': 9248,
   'name': 'G protein-coupled receptor 50',
   'symbol': 'GPR50',
   'taxid': 9606},
  {'_id': 'ENSG00000234696',
   '_score': 5.1264653,
   'name': 'GPR50 antisense RNA 1',
   'symbol': 'GPR50-AS1',
   'taxid': 9606},
  {'_id': 'ENSG00000269993',
   '_score': 5.111959,
   'symbol': 'AF003625.3',
   'taxid': 9606}],
 'max_score': 5.1352987,
 'took': 1202,
 'total': 6}

• A new field genomic_pos_hg19 is added to hold the genomic location data based on hg19:

In [4]: mg.getgene('1017', fields='genomic_pos_hg19')  
Out[4]:  
{'_id': '1017',
 'genomic_pos_hg19': {'chr': '12',
  'end': 56366568,
  'start': 56360553,
  'strand': 1}}

• A new field exons_hg19 is added to hold the exons data based on hg19:

In [5]: mg.getgene('1017', fields='exons_hg19')  
Out[5]:  
{'_id': '1017',
 'exons_hg19': {'NM_001290230': {'cdsend': 56365409,
   'cdsstart': 56360792,
   'chr': '12',
   'exons': [[56360552, 56360908],
    [56361832, 56361953],
    [56362561, 56362732],
    [56364827, 56365031],
    [56365304, 56366573]],
   'strand': 1,
   'txend': 56366573,
   'txstart': 56360552},
  'NM_001798': {'cdsend': 56365409,
   'cdsstart': 56360792,
   'chr': '12',
   'exons': [[56360552, 56360908],
    [56361640, 56361718],
    [56361832, 56361953],
    [56362561, 56362732],
    [56363258, 56363360],
    [56364827, 56365031],
    [56365304, 56366573]],
   'strand': 1,
   'txend': 56366573,
   'txstart': 56360552},
  'NM_052827': {'cdsend': 56365409,
   'cdsstart': 56360792,
   'chr': '12',
   'exons': [[56360552, 56360908],
    [56361640, 56361718],
    [56361832, 56361953],
    [56362561, 56362732],
    [56364827, 56365031],
    [56365304, 56366573]],
   'strand': 1,
   'txend': 56366573,
   'txstart': 56360552}}}

• You can still make Genome interval query based on hg19 by adding a hg19. prefix:

In [6]: mg.query('hg19.chrX:151,073,054-151,383,976', species='human')  
Out[6]:  
{'hits': [{'_id': 'ENSG00000231937',
   '_score': 6.9943757,
   'symbol': 'RP11-329E24.6',
   'taxid': 9606},
  {'_id': '574412',
   '_score': 6.9943757,
   'entrezgene': 574412,
   'name': 'microRNA 452',
   'symbol': 'MIR452',
   'taxid': 9606},
  {'_id': 'ENSG00000228965',
   '_score': 6.9943757,
   'symbol': 'RP11-1007I13.2',
   'taxid': 9606},
  {'_id': '2564',
   '_score': 6.9620624,
   'entrezgene': 2564,
   'name': 'gamma-aminobutyric acid (GABA) A receptor, epsilon',
   'symbol': 'GABRE',
   'taxid': 9606},
  {'_id': '4109',
   '_score': 6.9620624,
   'entrezgene': 4109,
   'name': 'melanoma antigen family A, 10',
   'symbol': 'MAGEA10',
   'taxid': 9606},
  {'_id': '407009',
   '_score': 6.9609237,
   'entrezgene': 407009,
   'name': 'microRNA 224',
   'symbol': 'MIR224',
   'taxid': 9606},
  {'_id': 'ENSG00000229967',
   '_score': 6.9609237,
   'symbol': 'RP11-366F6.2',
   'taxid': 9606},
  {'_id': 'ENSG00000266560',
   '_score': 6.9609237,
   'symbol': 'RP11-1007I13.4',
   'taxid': 9606},
  {'_id': '2556',
   '_score': 6.8982496,
   'entrezgene': 2556,
   'name': 'gamma-aminobutyric acid (GABA) A receptor, alpha 3',
   'symbol': 'GABRA3',
   'taxid': 9606},
  {'_id': '4103',
   '_score': 6.8982496,
   'entrezgene': 4103,
   'name': 'melanoma antigen family A, 4',
   'symbol': 'MAGEA4',
   'taxid': 9606}],
 'max_score': 6.9943757,
 'took': 1095,
 'total': 12}

As a final note, this change affects human genes only, of course.

Over the past several months, I have been working on the development of this software to allow users of R to incorporate MyGene.info's superb web services into their analyses. What actually began as a coding practice side project, became a full fledged effort to contribute to Bioconductor's repository.

Recently, I was able to present the work in progress at BioC2014, Bioconductor's annual conference that highlights developments and breakthroughs in genomic analyses in Boston. After a short talk and poster session, I received imperative feedback from staff as well as other scientists who use R/Bioconductor that I believe helped guarantee its acceptance.

Aside from the discovery how collaborative the community of R is, which allowed me to tailor the code and functions towards ease of integration within other Bioconductor packages, I learned how elaborate and standardized the documentation must be. So any ambiguity should be completely relieved by the manual pages and vignettes (R CMD check/build wouldn't cut me any slack!). mygene is completely free and open-source. Give it a try by downloading and installing from Bioconductor or install from your R console:

source("http://bioconductor.org/biocLite.R")  
biocLite("mygene")  

You must first upgrade to the latest Bioconductor version if you have not:

biocLite("BiocUpgrade")  

Here are a few quick examples of code for using mygene.
The user is exposed to four primary functions for getting annotations: getGene, getGenes, query, and queryMany. Use getGene, the wrapper for GET query of ”/gene/” service, to return the gene object for the given geneid.

> gene <- getGene("1017", fields="all")
> length(gene)
[1] 36
> gene$name
[1] "cyclin-dependent kinase 2"
> gene$taxid
[1] 9606
> gene$uniprot
$`Swiss-Prot` [1] "P24941"
$TrEMBL
[1] "B4DDL9" "E7ESI2" "G3V317" "G3V5T9"

Just as easily you can retrieve annotations from several thousand genes without worrying about clogging up our servers. Use queryMany, a wrapper for POST query of ”/query”service, to return the batch query result.

> xli <-c('DDX26B','CCDC83','MAST3','FLOT1','RPL11','ZDHHC20','LUC7L3','SNORD49A','CTSH')
> queryMany(xli, scopes="symbol", fields="entrezgene", species="human")
Finished  
DataFrame with 9 rows and 3 columns  
query entrezgene         _id  
<character>  <integer> <character>  
1      DDX26B     203522      203522  
2      CCDC83     220047      220047  
3       MAST3      23031       23031  
4       RPL11       6135        6135  
5     ZDHHC20     253832      253832  
6      LUC7L3      51747       51747  
7    SNORD49A      26800       26800  
8        CTSH       1512        1512  

Two utility functions are also available: metadata, and makeTxDbFromMyGene. metadata allows the user to retrieve MyGene.info’s metadata and view available fields for querying.

> mg<-MyGene()
> metadata(mg)
> metadata(mg)$available_fields
[1] "accession"         "alias"             "biocarta"          "chr"               "end"              
[6] "ensemblgene"       "ensemblprotein"    "ensembltranscript" "entrezgene"        "exons"            
[11] "flybase"           "generif"           "go"                "hgnc"              "homologene"       
[16] "hprd"              "humancyc"          "interpro"          "ipi"               "kegg"             
[21] "mgi"               "mim"               "mirbase"           "mousecyc"          "name"             
[26] "netpath"           "pdb"               "pfam"              "pharmgkb"          "pid"              
[31] "pir"               "prosite"           "ratmap"            "reactome"          "reagent"          
[36] "refseq"            "reporter"          "retired"           "rgd"               "smpdb"            
[41] "start"             "strand"            "summary"           "symbol"            "tair"             
[46] "taxid"             "type_of_gene"      "unigene"           "uniprot"           "wikipathways"     
[51] "wormbase"          "xenbase"           "yeastcyc"          "zfin"   

makeTxDbFromMyGene allows the user to store transcript annotations from a mygene 'exons' query in a sqlite database. TxDb is a really convenient Bioconductor container that allows for interoperability with other packages for utilization and accession of transcript annotations.

> txdb <- makeTxDbFromMyGene(xli, scopes="symbol", species="human")
> transcripts(txdb)
GRanges object with 22 ranges and 2 metadata columns:  
seqnames                 ranges strand   |     tx_id      tx_name  
<Rle>              <IRanges>  <Rle>   | <integer>  <character>  
[1]           1   [24018268, 24022915]      +   |        13 NM_001199802
[2]           1   [24018268, 24022915]      +   |        14    NM_000975
[3]          11   [85566143, 85631063]      +   |         2 NM_001286159
[4]          11   [85566143, 85631063]      +   |         3    NM_173556
[5]          13   [21946709, 22033508]      -   |        18    NM_153251
...         ...                    ...    ... ...       ...          ...
[18] 6_mann_hap4 [  2043604,   2058547]      -   |         9    NM_005803
[19]  6_mcf_hap5 [  2077393,   2090996]      -   |        10    NM_005803
[20]  6_qbl_hap6 [  1988442,   2003382]      -   |         5    NM_005803
[21] 6_ssto_hap7 [  2027817,   2042770]      -   |         7    NM_005803
[22]           X [134654554, 134716460]      +   |         1    NM_182540
-------
seqinfo: 15 sequences from an unspecified genome; no seqlengths  

Feedback is very welcome and appreciated. Thanks to Ryan Thompson, and the Su lab, especially Chunlei and Tobias for the guidance!

To install:

source("http://bioconductor.org/biocLite.R")  
biocLite("mygene")  

If you have an older version of Bioconductor installed, upgrade it first:

biocLite("BiocUpgrade")  

Your feedback is welcome!