Biological sequence analysis and information processing by artificial neural networks

1
Biological sequence analysis and information
processing by artificial neural networks

• Søren Brunak
• Center for Biological Sequence Analysis
• Technical University of Denmark
• brunak_at_cbs.dtu.dk

2
Pairwise alignment

• gtcarp Cyprinus carpio growth hormone
  210 aa vs.
• gtchicken Gallus gallus growth hormone
  216 aa
• scoring matrix BLOSUM50, gap penalties -12/-2
• 40.6 identity Global alignment
  score 487
• 10 20 30
  40 50 60 70
• carp MA--RVLVLLSVVLVSLLVNQGRASDN-----QRLFNNAVIR
  VQHLHQLAAKMINDFEDSLLPEERRQLSKIFPLSFCNSD
• . .... . . .
  .. . .. . ... . . .
• chicken MAPGSWFSPLLIAVVTLGLPQEAAATFPAMPLSNLFANAVLR
  AQHLHLLAAETYKEFERTYIPEDQRYTNKNSQAAFCYSE
• 10 20 30 40
  50 60 70 80
• 80 90 100 110
  120 130 140 150
• carp YIEAPAGKDETQKSSMLKLLRISFHLIESWEFPSQSLSGTVS
  NSLTVGNPNQLTEKLADLKMGISVLIQACLDGQPNMDDN
• ..... .... . .
  ... . ... . .... . .
• chicken TIPAPTGKDDAQQKSDMELLRFSLVLIQSWLTPVQYLSKVFT
  NNLVFGTSDRVFEKLKDLEEGIQALMRELEDRSPR---G
• 90 100 110 120
  130 140 150 160
• 170 180 190
  200 210

3
(No Transcript)
4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
Biological neuron
12
(No Transcript)
13
Diversity of interactions in a network enables
complex calculations

• Similar in biological and artificial systems
• Excitatory () and inhibitory (-) relations
• between compute units

14
(No Transcript)
15
Transfer of biological principles to neural
network algorithms

• Non-linear relation between input and output
• Massively parallel information processing
• Data-driven construction of algorithms
• Ability to generalize to new data items

16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
Simplest non-trivial classification problem

• CNHSYYP, HIETRRA, NWQSADY, NQYSEPR, WHITRCA,
  DYHSANY, ...
• Two categories positives and negatives
• Data described by two features, e.g.
• charge, sidechain volume, molecular
• weight, number of atoms, ...

20
Features of phosphorylations sites
PKG cGMP- dep.kinase
cdc2 Cyclin- dep.kinase 2
CK-II Casein kinase 2
PKC
CaM-II Ca/cal-modulin-dep. kinase
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
Homotypical cerebral cortex (from primate) - 6
layers
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
DEMO
30
(No Transcript)
31
(No Transcript)
32
Training and error reduction
negative
positive
33
Transfer of biological principles to neural
network algorithms

• Non-linear relation between input and output
• Massively parallel information processing
• Data-driven construction of algorithms

34
(No Transcript)
35
Sparse encoding of amino acid sequence windows
36
Sparse encoding of nucleotide sequence windows
Nucleotides 4 letter alphabet Normally no need
for a fifth letter ACGTAGGCAATCTCAGACGTTTATC 10
00010000100001100000100010010010001000000101000001
010010000010100001000010000100010001100000010100
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)