The UniProt Metal Binding Site Machine Learning Challenge

 We would like to invite the machine learning community to help UniProt by creating 

computational methods to predict metal binding sites across the whole of UniProtKB. 

At present around 17% of curated proteins have annotated metal binding site residues, 

which our curators have carefully identified from the literature or known structures 

from PDB. UniProt identifies the specific amino acid residues that participate in metal

binding sites and also which metal is bound. For example, for the Neurospora crassa

metallothionein protein (shown below) contains 7 cysteine residues involved in binding 

6 copper ions.

 

  

When we look at the uncurated TrEMBL section which contains the large majority of known 

protein sequences we see that just 3% of proteins have an annotated metal binding site. 

These annotations are created by a variety of automated annotation methods currently 

used. The difference in coverage between the reviewed (Swiss-Prot) and unreviewed 

(TrEMBL) suggests that there are many millions of missing metal binding site annotations 

in the 225 million TrEMBL sequences.

We would like to invite interested researchers to take part in a challenge to create new 

methods to rapidly predict metal binding site annotations that can be deployed by UniProt

as part of its automatic annotation pipeline. These methods could be completely based on 

sequence data or perhaps incorporate information from known and/or predicted structures. 

Although we don’t want to prejudge what methodology may work, we are particularly keen 

that methods be both accurate and very fast for scalability. All data and software must be 

open and not under restrictive licensing terms.

If you would like to take part in this initiative please register your interest by filling out

 this google form by 18th February 2022. We will then hold a planning meeting with 

the participants to discuss timelines and evaluation of the methods.

 

Navigating the use of tools on the new UniProt website

                                                 (https://beta.uniprot.org/)

One of the biggest changes we have made on the new UniProt website is to improve your

 usage of the tools associated with our data - BLAST, Align, ID mapping and Peptide 

mapping.

You can easily access all of these from the homepage of the new website,

 

 

 

 

 

 or from the toolbar of any entry.

 

When you open BLAST, you can search by on any UniProtKB ID (e.g. P05067 or 

A4_HUMAN) or a UniParc protein sequence archive ID (e.g. UPI0000000001) to

autofill the protein sequence. One of the biggest changes we have made is to allow 

you to select the species you wish to BLAST against, so if you are only interested in

dog (TaxID:9615), start to type the common (dog) or species (Canis lupus familiaris) 

name and the autocomplete will help you select the correct taxonomic group 

(TaxID:9615), to search for a match. You can choose to limit your search to any 

level of the taxonomic tree, so you can search your dog sequence against, for example,

all mammalian sequences by using the appropriate taxonomic identifier (TaxID:40674).

You can also upload a list of sequences to BLAST, up to a maximum of 20, from an 

external file. You can reset the BLAST parameters, should you wish, and name your 

job so you can easily identify when you return in up to a week's time.

Once your job has run,, it will appear in the new Tool Results view. Here you can see 

all of your  jobs that have completed over the last 7 days, identified by the names you 

have chosen to give them. 

When you open your job, we have increased the number of ways you can view the results.

 For example, in the case of a Sequence Alignment, you can take a look at the traditional 

view of the aligned amino acids, with a summary of any one of the entries you select acting 

as a reference sequence below, or you can see a phylogenetic view or a percent identity 

matrix.

Loading data from a file (up to a maximum of 100 sequences) and restricting the search

criteria by species are features also available in the Peptide Search tool, where users, 

for example proteomics scientists, can match peptide to protein, and an increasing 

number of database identifiers are now available in the ID mapping tool for users to 

switch between.

We hope you find this first look at the new beta version of the UniProt website useful, 

and easy to navigate. Please let us know if you have any problems at all, or if you can 

think of any way in which it can be further improved by using the contact form accessible 

from every page.

A New Look to UniProt

 https://beta.uniprot.org/

 

The UniProt Consortium is happy to announce a new look and feel to our web pages, designed to improve the user experience and enhance your journey of discovery through the world of protein science. We have only released the beta version of the site so far, but please take time to explore and send us your feedback - we will be working hard over the next few months to improve the site, ready for it to replace our current website in 2022.

Here are a few examples to get you started, but please make use of the entire website and let us know where we can still make further improvements.

You can find your favorite protein with a simple search - for example, by typing ‘hemoglobin’ or a function description 'apoptosis' or even a disease name such as ‘cystic fibrosis’ in the search box.

 

This brings up many hits in our new ‘card view’ but it is easy to filter your search results down to more manageable numbers by selecting a species or by limiting the results to reviewed, expertly curated entries using the left-hand menu.

The card view gives a quick overview of what data you can find in each entry - the CFTR protein has 210 reviewed variants and 38 3D structures, for example. The entries are ranked by the amount of content, so you will always see those with the most information at the top of your search.

We have embedded our ProtVista feature viewer directly into the entry so that you can both see and read about protein sequence features at the same time.

You can also see the structured Gene Ontology annotation in a summarized ribbon view, developed by the Alliance for Genomic Resources.

You can view protein 3D structures - both experimental (when available) or as predicted by AlphaFold (http://insideuniprot.blogspot.com/2021/07/).

And take a look at the subcellular location of your protein, pictured using SwissBioPics.

 

We hope you find this first look at the new UniProt website useful, and easy to navigate. Please let us know if you have any problems at all, or if you can think of any way in which it can be further improved by using the contact form accessible from every page.

AlphaFold structure predictions freely available in UniProt

Today marks a historic moment in the world of molecular structural research. AlphaFold structure predictions covering the proteomes of 21 species have been made freely available to the scientific community, and we are delighted to announce that they are also accessible through the UniProt website and the Protvista protein viewer. AlphaFold is an Artificial Intelligence (AI) system developed by DeepMind that predicts the three-dimensional (3D) structure from an amino-acid sequence.

Currently, the Protein Data Bank (PDB) contains over 180,000 macromolecular structures which cover ~55,000 UniProt Knowledgebase (UniProtKB) proteins. There are ~7,300 human UniProt proteins with PDB structures and AlphaFold predicts structures for 20,610. This covers most of the human reference proteome! With more than 220 million proteins in UniProtKB today, AlphaFold represents a great opportunity to predict structures for millions of proteins in thousands of species. The availability of large numbers of predicted models can provide important insights on the function of proteins and their role in the species biology. They can be used to kick-start experimental de novo structure determination.

In November 2020, AlphaFold was recognised as the best-performing method for predicting 3D protein structure by the assessors of the 14th CASP experiment. The best-predicted 95% of residues in AlphaFold models had a median alpha carbon RMSD of 0.96 Å to the experimental models, compared to 2.83 Å for the next-best method. Thus, the AlphaFold predictions were very similar to the experimentally determined structures of the proteins included in this round of CASP. The AlphaFold prediction also includes a predicted model-quality score for individual residues indicating regions of high quality and those where the model is probably less reliable. 

And the future is looking bright. In the coming months AlphaFold will be expanded to cover a large proportion of all catalogued proteins in the UniProt, UniRef90 clusters. This means that for every known sequence in the UniProt data resource there will be either an experimentally determined structure in PDB, or an AlphaFold model in the AlphaFold database. This development represents a step-change for molecular biology - for the first time in history, for almost every protein of known sequence, a high-quality 3D model will be readily available. 

Open data resources like PDB and UniProt have been key in the successful development of AlphaFold. AlphaFold structure predictions freely accessible to the scientific community is a success story for academia-industry collaboration and for open science.