Sense-Antisense Proteins

1
Sense-Antisense Proteins
Vision Lab Presentation Ruchir Shah April 16,
2003
2
Sense-Antisense Proteins
Peptides generated from sense and antisense DNA
strands have inverted hydropathies. Although
it makes no sense, it is hypothesized that S- and
AS-peptides could have a high binding affinity
for each other.
Picture adapted from J.R.Heal et al ChemBioChem
2002,3,136-151.
3
S-AS Codon Table
4
Inverted Hydropathy
BlueNon Polar PinkPolar
Picture adapted from J.R.Heal et al ChemBioChem
2002,3,136-151.
5
S-AS Codons

• Degeneracy
• One sense AA can have more than One antisense
  AA.
• Hydropathy
• Sense antisense AAs have inverted hydropathy.
• Codon biases/codon frequencies?
• Sense proteins interact with Antisense proteins
• Numerous experimental evidences suggest that
  Sense and
• AS peptide have specific binding Affinity.

6
Experimental evidences
Picture adapted from J.R.Heal et al ChemBioChem
2002,3,136-151.
7
How do S-AS Amino Acids interact?
Picture adapted from J.R.Heal et al ChemBioChem
2002,3,136-151.
8
Molecular Recognition Theory
Picture adapted from J.R.Heal et al ChemBioChem
2002,3,136-151.
9
Tasks

• Literature says
• S-AS proteins exist
• S-AS proteins interact specifically with each
  other!
• Tasks
• Look for S-AS protein pairs(how such many pairs
  exist?)
• What are the biological implications?
• Do they really interact?

10
How to find S-AS pairs from Sequence Db?

• Conventional Sequence identity tools can be used
  to find out similar proteins.
• Example
• Blast or Smith Waterman with a choice of
  substitution matrix
• Positive score for Identity or desirable
  substitutions.
• Negative score for undesirable substitutions.

11
BLOSUM 62
Source http//www.blc.arizona.edu/courses/bioinfo
rmatics/blosum.html
12
Design of a new substitution matrix

• To find S-AS pairs using existing sequence
  identity tools I need a special matrix.
• New matrix should
• - positively score S-AS pairs
• - negatively score other pairs
• - reflect the degeneracy of genetic code
• - average score should be negative to avoid
  false positives!!

13
S-AS Codon Table
14
(No Transcript)
15
Results What does it look like?
It works!!
16
Results contd..
Low complexity regions!
17
Lots of small hits(lessons learnt!) get rid of
noise/background get rid of Low complexity
regions use a better matrix
18
Design of a new substitution matrix

• New matrix should
• - positively score S-AS pairs
• - negatively score other pairs
• - reflect the degeneracy of genetic code
• -take into account the codon biases

19
S-AS Codon Table
Source SGD(Stanford) Saccharomyces Genome Databas
e
20

1. Low complexity filter SEG
2. More meaningful Matrix Formula for new scoring
   scheme

21
Flow Chart
Sequence database (Yeast) 6000prtns
Run Smith Waterman All against All With new matrix
Look for hits
Compare it with Interaction data
22
Tasks

• Look for sense-antisense protein pairs
• in protein sequence databases.
• List all sense-antisense pairs
• Compare the list with List of interacting
• proteins.
• Example

Sense-Antisense pairs
Database of Interacting Prtns
P5-P99 P2-P102 P104-P4
P1-P101 P2-P102 P3-P103 P4-P104
23
Tasks

• Look for sense-antisense protein pairs
• in protein sequence databases.
• List all sense-antisense pairs
• Compare the list with List of interacting
• proteins.
• Example

Sense-Antisense pairs
Database of Interacting Prtns
P5-P99 P2-P102 P104-P4
P1-P101 P2-P102 P3-P103 P4-P104
24
DIP Database of Interacting Proteins http//dip.
doe-mbi.ucla.edu/dip/Main.cgi
SSsmall scale experiment HThigh throughput
exp. Purpleoverlap Bars more than 1 exp.
Proteins 4727 Interactions 15174
25
Work in Progress

• Statistics of alignment
• Distinguish random from meaningful hits!
• Relative entropy of the matrix
• Gap Penalties

26
Acknowledgments
Todd Vision (Biology) Alex Tropsha (Pharmacy) Dr.
Falk (Nephrology) All of my lab mates.