Modeling virus-host interaction networks

1
Modeling virus-host interaction networks

• Perry Evans
• Mahdi Sarmadi

2
NEF interaction sites
Roeth and Collins MMBR 06?
3
HIV-1 ELMs
4
Interaction prediction

• Identify ELMs on HIV-1 proteins
• Find human proteins with functional CDs
• Validate predictions using NIAIDs HIV, human
  interaction database of 2K experimental
  interactions

H1
H2
5
Global validation
6
HIV-1 vs. HCV ELMs
4 interesting unique ELMs
7 interesting unique ELMs
7
MAPK study

•  

8
Hub modularity predicts interactions

•  

9
De Novo Motif Discovery

• Known motifs do not unveil the grammar of
  interaction between HIV-1 and human proteins
• Computational motif discovery of de novo motifs
  is based on the virtue of Over-Representation

10
Over Representation
11
HIV-1 Protein Alignment Contains Invariant Regions
HIV-1 TAT protein (first 65 AA of the sequence is
shown)
12
H2 Motif Finding Results
Motif Pattern
HIV
H1 Protein
H2 Proteins
H2 Matches
Occ. On HIV
HPRD P-Val
NCBI P-Val
FWY.0,1C..C
TAT
EP300 (E1A binding protein p300)
207
41
100
1.45E-04
1.04E-05
GTF2H1 (General Transcription factor IIH)
26
5
5.25E-02
3.15E-02
TBP (TATA box binding protein)
83
15
1.76E-02
5.94E-03
TP53 (Tumor protein p53)
238
42
8.10E-04
6.61E-05
TP73 (Tumor protein p73 )
31
8
4.19E-03
1.83E-03
CREBBP (CREB binding protein)
199
32
1.94E-02
3.69E-03
KRKR
96
94
8.89E-10
1.53E-10
ILKQE
VPR
20
50
1.02E-07
7.46E-11
13
H2 Set Motif Alignment
14
FWY.0,1C..C on TAT
15
(No Transcript)
16
H2 Motifs Counter Motif?
17
H2 Motifs Counter Motif?
E.FIV..IL.K
18
HIV Invariant Region Motifs