Structural Immunoinformatics

1
Structural Immunoinformatics two case studies
Medical University of Sofia Faculty of Pharmacy

• M. Atanasova, I. Dimitrov, A. Patronov, I.
  Doytchinova

Regional Conference in Supercomputing
Applications in Science and Industry, 20-21
Sept. 2011, Sunny Beach, Bulgaria
2
Medical University of Sofia Faculty of Pharmacy
Immunoinformatics (Computational Immunology,
Theoretical Immunology)
Immunoinformastic Approaches
Structure - based
Sequence - based
peptide pIC50exp ILDPFPVTV 8.654 ALDPFP
PTV 8.170 VLDPFPITV 8.139 ................ .......
FLDPFPATV 8.270
Affinity f (Interaction energy) Molecular
docking Molecular dynamics
Affinity f (Chemical Structure) Motif-based,
QMs, ANN, SVM
3
Medical University of Sofia Faculty of Pharmacy
Case study 1 T-cell epitope prediction of
proteins from Boophilus microplus

• Ticks are hematophagous parasites that feed on
  variety of domestic animals.
• B. microplus tick
• a hard tick
• transmits lethal pathogens
• causes disease and death.

Boophilus microplus tick.
Aim to predict peptides originating from B.
microplus and binding with high affinity to
murine MHC class II proteins IAd and IEd.
Collaborator University of Pretoria, SA
4
Medical University of Sofia Faculty of Pharmacy
Case study 1 T-cell epitope prediction of
proteins from Boophilus microplus
MHCPred and RANKPEP servers for MHC class II
binding prediction
Sequence - based
Approaches used
Structure - based
Molecular docking calculations
5
Medical University of Sofia Faculty of Pharmacy
Case study 1 T-cell epitope prediction of
proteins from Boophilus microplus
Workflow
B. Microplus number Protein
peptides Contig2828 59 Contig7420 93 CK181624
61
Selection of high immunogenic B. microplus
proteins by VaxiJen server.
Presentation of the selected proteins as sets of
overlapping peptides.
1.
Selection of input X-ray structures of complexes
of murine MHC II protein with a peptide.
Homology modeling of IEk to IEd structure
2.
Ova/IAd (pdb code 2iad) HB/IEk (pdb code1iea)
Optimization of complexes of each peptide with
each MHC II protein.
Docking calculations
3.
biding site - 6Å Chemscore scoring function
fixed protein and peptide backbone ranking by
score GOLD v.5.0.2.
6
Medical University of Sofia Faculty of Pharmacy
Case study 1 T-cell epitope prediction of
proteins from Boophilus microplus
Binding affinity prediction to IAd by MHCPred and
RANKPEP
MHCPred Predicted binders with IC50 lt 50 nM are
highlighted in green.
RANKPEP Predicted binders with binding
threshold 7.10 are highlighted in purple.
7
Medical University of Sofia Faculty of Pharmacy
Case study 1 T-cell epitope prediction of
proteins from Boophilus microplus
Binding affinity prediction to IAd and IEd by
Molecular docking
IAd
IEd
The top 2 best clusters of binders are
highlighted in magenta.
8
Medical University of Sofia Faculty of Pharmacy
Case study 1 T-cell epitope prediction of
proteins from Boophilus microplus
Peptides selected for further experimental
studies
9
Medical University of Sofia Faculty of Pharmacy
Case study 2 Prediction of peptide binding to
Swine Leukocyte Antigen (SLA-1)

• Swine Influenza in pigs
• An acute respiratory disease
• High morbidity depending on the
• immune status
• Can results in important economic
• losses.

Aim to generate quatitative matrices (QMs) for
prediction of peptides binding to SLA-1
CReSA Centre de Recerca en Sanitat Animal
10
Medical University of Sofia Faculty of Pharmacy
Case study 2 Prediction of peptide binding to
Swine Leukocyte Antigen (SLA-1)

• Workflow
• Homology modeling of SLA-1
• from HLA0201 (pdb3pwj).
• 2. Construction of combinatorial
• library of peptides.
• 3. Molecular docking of peptides
• to SLA proteins.
• 4. Forming of docking score-based

Modeled proteins SLA-10101
SLA-10401 SLA-10501 SLA-11101 7
anchor positions X 19 aa 133 1 original
ligand 134 peptides biding site - 6Å
Chemscore scoring function fixed protein and
ligand apart from the residues from the tested
peptide position ranking by lowest RMS GOLD
v.5.0.2. normalization of the binding energies
and compilation into QMs.
11
Medical University of Sofia Faculty of Pharmacy
Case study 2 Prediction of peptide binding to
Swine Leukocyte Antigen (SLA-1)

• Workflow
• Homology modeling of SLA-1
• from HLA0201 (pdb3pwj).
• 2. Construction of combinatorial
• library of peptides.
• 3. Molecular docking of peptides
• to SLA proteins.
• 4. Forming of docking score-based

12
Medical University of Sofia Faculty of Pharmacy
Case study 2 Prediction of peptide binding to
Swine Leukocyte Antigen (SLA-1)
SLA allele Pocket 2 profile P2 accepts
0101 Tyr9, Ala24, Val34, Met45, Glu63, Lys66, Gln67 Leu, Met, Asn
0501 Ser9, Ala24, Val34, Met45, Glu63, Lys66, Gln67 Trp, Leu, Phe
0401 Tyr9, Ala24, Val34, Met45, Glu63, Asn66, Val67 Leu, Met, Thr
1101 Ser9, Glu24, Val34, Met45, Glu63, Arg66, Val67 Leu, Met, Ile
13
Medical University of Sofia Faculty of Pharmacy
Case study 2 Prediction of peptide binding to
Swine Leukocyte Antigen (SLA-1)
SLA allele Pocket 2 profile P2 accepts
0101 Tyr9, Ala24, Val34, Met45, Glu63, Lys66, Gln67 Leu, Met, Asn
0501 Ser9, Ala24, Val34, Met45, Glu63, Lys66, Gln67 Trp, Leu, Phe
0401 Tyr9, Ala24, Val34, Met45, Glu63, Asn66, Val67 Leu, Met, Thr
1101 Ser9, Glu24, Val34, Met45, Glu63, Arg66, Val67 Leu, Met, Ile
14
Medical University of Sofia Faculty of Pharmacy
Case study 2 Prediction of peptide binding to
Swine Leukocyte Antigen (SLA-1)
SIV proteins screened to predict SLA binders -
hemagglutinin (HA) - nucleocapsid protein (NP) -
matrix protein 1 (M1) - polymerase PB1 (PB1)
15
Thank you for your attention!