Computational Modelling Group

Dr. Alistair Bailey

Research Fellow
Medicine (FM)
Complete this online contact form to contact Alistair.

My background in in engineering and I am now part of an interdisciplinary team working across the fields of immunology, structural biology, systems biology and bioinformatics.

In Southampton, I am member of the Cancer Sciences Unit, the Centre for Proteomic Research, and the Institute for Life Sciences. I also have ongoing collaborations with the Biological NMR Group ( with the Biological Computation group at Microsoft Research Cambridge.

Research Interests

My research is primarily into the MHC class I antigen processing and presentation pathway.

Major histocompatibility complex class I molecules (MHC I) present small protein fragments called peptides to our immune systems at the surface of almost all nucleated cells. The function of MHC I molecules is to select peptides from a large intracellular pool that best create stable molecules and they are assisted in this process by co-factor molecules, notably tapasin.

This is the key mechanism by which our cells inform our immune systems as to whether they are healthy or have become cancerous or become infected by pathogens such as viruses or bacteria.

Our ultimate goal is to produce a predictive, quantitative models of this peptide selection process by linking together observations based upon molecular structures with observations of the behaviour of these molecules as part of a system leading to the immune response. We do this through an iterative process of modelling and laboratory experiments.

  • At a structural level I am using bioinformatics and molecular dynamics as a computational microscope to identify links between protein dynamics and protein function.

  • A systems biology approach has been developed in conjunction with Microsoft Research, Cambridge. This uses a mathematical model of the kinetic processes to describe the interaction of these proteins and quantify peptide presentation as observed experimentally.

Working with...

Timothy Elliott
Professor, Medicine (FM)
Jorn Werner
Reader, Biological Sciences (FNES)



MHC Class I docking with Tapasin:

Tapasin is depicted in green and MHC Class I allele HLA-B*0801 is depicted with the heavy chain in blue and red and the B2m sub-unit in grey:

(A) The docking of Tapasin and MHC is driven by biochemical interaction information. The TN6 region of Tapasin identified by Dong et al. is docked using HADDOCK with residue T134 of MHC. T134 has been identi fied as key to Tapasin-MHC interaction. An open structure of MHC Class I allele HLA-B0801 (pdb ID: 1agb) is generated by molecular dynamic simulation. The human Tapasin structure is completed from the crystal structure (pdb ID: 38fu) using homology modelling.

(B) Tapasin residue R187 of the TN6 region nds a favourable interaction with T134. The side chain of T134 of the peptide-bound crystal Structure of HLA-B*0801 is depicted in gold and the docked structures side chains are shown in red, white and blue. Movement of the alpha2-1 helix (red) widens the F-pocket at the peptide C-terminus by about 3 Angstroms as previously observed by Sieker et al. This movement of the alpha2-1 helix corresponds with a movement of T134 by about 2 Angstroms from the closed structure to nd its interaction with R187 on Tapasin in this simulation. E185 on Tapasin appears to form a structural interaction with R187.

(C) Colours as panels B. The simulation independently identifi es a favourable C-terminus interaction between Tapasin residue R333 and E222 on MHC. An E222K mutation of murine MHC allele H2-Dd has been shown to prevent binding to Tapasin.

A putative model of the Peptide Loading Complex:

A simulated peptide free MHC Class I- 2m dimer HLA-B*0801 (yellow) is docked to Tapasin (green) using HADDOCK.

ERp57 (purple) forms a conjugate with Tapasin.

A model structure of Calreticulin (orange) is positioned approximately with the glycan binding sites close to that of MHC I and the proline domain tip close to the domain residues of ERp57 with which it is believed to interact.

TAP is not shown as no structure is available.

MHC CLass I Structure:

(a) A Van der Waals surface representation of lumenal domains of HLA-B*0801 heterodimer.

(b) A ribbon representation indicating Ig-type 3 domain, 1 and 2 helices and sheets create the peptide binding groove. The alpha-2 helix is thought to be where Tapasin interacts. Ig-type protein B 2m is noncovalently bound, but provides rigidity to the protein.

(c) Plan view of the peptide binding groove with a bound peptide coloured orange. It is the most polymorphic part of MHC Class I proteins, where 8-10 residue peptides bind forming complexes for presentation to CD8+ T-Cells. Pockets identified as anchors for the peptide are annotated B in yellow, C in dark blue, D in pink and F in red.

(d) A cartoon representation of MHC Class I structure indicating the transmembrane nature of the protein with the C-terminus tail that extends across membranes which is not seen in the lumenal domains comprising the crystal structures shown in (a) or (b).

A kinetic model of peptide editing:

In this graphical representation the yellow shapes represent MHC I- B2m dimers, the red shapes are peptides and the green shapes represent Tapasin.

Each box represents a reaction. A reversible reaction box contains two rates with the forward reaction rate on top and the backward reaction rate below.

Irreversible reactions contains only one reaction rate.

Incoming arrows represent reactants and outgoing arrows represent products.

Crossed circles represent generation or degradation of a protein species.

HLA-B*0801 molecular dynamics simulation using Gromacs

Click on the image to start the animation.

Project Map showing all the people, techniques and alleles involved

MHC Class I antigen processing schematic