10A should also be red

same as 10B

Just as Dr. Kister said, when black SM has odd number of strandons, there's something wrong.
Now we know, SOMETHING=RED.

why 2AF not exist?

2AF follows every rules, why there's no 2AF?

motifs in SM contents odd number strandons are all red in some SM

ex: SM#7, SM#11, maybe SM#16

error correction status

error type	corrected	descibtion	example
motif not exist	yes	auto fix	2A, 2X, 2AL, 3C, 3J, 3N, 3AB, 4N, 23E
jump caused by missing strand	yes	add'	1Y, 2S, 2Y,2AM,3P,3Q,11A
other			2L: should delete strand 6
wrong direction caused by uninverted sheet	yes		3AB, should be 2F
assign multiple motif number for same motif			3P, 3Q
combinition of two motif			11A

error correction in statistics table

error type	corrected	descibtion	example
motif not exist	not yet		2A, 2X, 2AL, 3C
jump caused by missing strand	not yet		1Y, 2S, 2AM
other			2L: should delete strand 6
wrong direction caused by uninverted sheet			3AB, should be 2F
assign multiple motif number for same motif			3P, 3Q
combinition of two motif			11A

change the table

1ciy (13A)

1
3 4 11 6 9

2 5 10 7 8

change to

1 4
3 4 11 6 9

2 5 10 7 8

d & j only on edge

wrong strandon: wrong direction and jump only exist on edge strandon except 7H.
(Not including wrong SM here)

idea about parallel and anti-parallel

A B
we got same-sheet parallel (A) only if the loop is long enough.
otherwise if two consicutive strands next to each other in the same sheet can only be anti-parallel like B.

idea about strandon prediction

in this kind of motif should have a preference signal like this.

which shows strand 1 like strand2, and strand 2 like strand 1 and strand 5, etc

ref papers

this paper use some probes about 5~9 residues long to predict the structures.
it's like in the middle of homology modeling(lots of residues) and my previous finding which use 4 residues to predict loops.
they obtained the probes by statistics similar to HMM.

this paper use similar approach as mine, they try to predict Strand-Loop-Strand structure.
I think they got a pretty good result, and I might use their result as input.

this paper try to predict folding initiation sites. which might solve our problem by this way.

this paper predict long range contact by contact maps.
I think this might help a bit in our case.

structure prediction

This paper use rules for B-sheet motifs to predict protein structures.
Which is pretty similar to ours.

plan 050907

align fold 3, SM#2, four strands in interlock by HB info.

check SM which has odd number of strandons

ex: SM#7

7E 5 (2 1) (8 9)
(4 3) (6 7)
if you change the order of 8 and 9, then it's not in SM#7, it will become a even number strandons.

discover rules/associations by weka

 
=== Run information ===

Scheme:       weka.associations.Apriori -N 10 -T 0 -C 0.9 -D 0.05 -U 1.0 -M 0.1 -S -1.0
Relation:     predictstrandon2.txt-weka.filters.unsupervised.attribute.Remove-R12-19,25-weka.filters.unsupervised.attribute.Remove-R1,14-15,17-18-weka.filters.unsupervised.attribute.Remove-R14-weka.filters.unsupervised.attribute.Remove-R9-weka.filters.unsupervised.attribute.Remove-R13
Instances:    61
Attributes:   13
              seq
              res1
              res2
              res3
              res4
              ss
              ss1
              ss2
              ss4
              type
              type2
              typen
              intrastrdonn
=== Associator model (full training set) ===


Apriori
=======

Minimum support: 0.75
Minimum metric : 0.9
Number of cycles performed: 5

Generated sets of large itemsets:

Size of set of large itemsets L(1): 4

Size of set of large itemsets L(2): 5

Size of set of large itemsets L(3): 2

Best rules found:

 1. intrastrdonn=FALSE 55 ==> ss2=C 55    conf:(1)
 2. ss1=C 50 ==> ss2=C 50    conf:(1)
 3. ss4=C 49 ==> ss2=C 49    conf:(1)
 4. ss4=C intrastrdonn=FALSE 48 ==> ss2=C 48    conf:(1)
 5. ss1=C intrastrdonn=FALSE 48 ==> ss2=C 48    conf:(1)
 6. ss4=C 49 ==> ss2=C intrastrdonn=FALSE 48    conf:(0.98)
 7. ss2=C ss4=C 49 ==> intrastrdonn=FALSE 48    conf:(0.98)
 8. ss4=C 49 ==> intrastrdonn=FALSE 48    conf:(0.98)
 9. ss1=C 50 ==> ss2=C intrastrdonn=FALSE 48    conf:(0.96)
10. ss1=C ss2=C 50 ==> intrastrdonn=FALSE 48    conf:(0.96)

strandon prediction using weka, result1

=== Run information ===

Scheme:       weka.classifiers.trees.ADTree -B 10 -E -3
Relation:     predictstrandon2.txt-weka.filters.unsupervised.attribute.Remove-R2,7-weka.filters.unsupervised.attribute.Remove-R14-17-weka.filters.unsupervised.attribute.Remove-R16-17-weka.filters.unsupervised.attribute.Remove-R18
Instances:    61
Attributes:   17
              pos
              res1
              res2
              res3
              res4
              score
              s1
              s3
              s3
              s4
              s5
              s6
              s7
              typen
              total
              relpos
              intrastrandon
Test mode:    evaluate on training data

=== Classifier model (full training set) ===

Alternating decision tree:

: -1.04
|  (1)s6 < 0.291: -1.434
|  (1)s6 >= 0.291: 0.773
|  |  (3)score < 1.401: -0.731
|  |  (3)score >= 1.401: 0.706
|  |  |  (5)s3 < 2.062: 1.42
|  |  |  (5)s3 >= 2.062: -0.452
|  (2)relpos < 6.5: -0.737
|  |  (4)res3 = T: 0.427
|  |  (4)res3 != T: -0.819
|  (2)relpos >= 6.5: 0.658
|  (6)res1 = A: 0.254
|  (6)res1 != A: -0.452
|  |  (7)pos < 54: 0.146
|  |  (7)pos >= 54: -0.418
|  |  |  (8)typen = &12: 0.119
|  |  |  (8)typen != &12: -0.378
Legend: -ve = FALSE, +ve = TRUE
Tree size (total number of nodes): 25
Leaves (number of predictor nodes): 17

Time taken to build model: 0.02 seconds

=== Evaluation on training set ===
=== Summary ===

Correctly Classified Instances          61              100      %
Incorrectly Classified Instances         0                0      %
Kappa statistic                          1    
Mean absolute error                      0.0237
Root mean squared error                  0.0445
Relative absolute error                 12.6362 %
Root relative squared error             14.9366 %
Total Number of Instances               61    

=== Detailed Accuracy By Class ===

TP Rate   FP Rate   Precision   Recall  F-Measure   Class
  1         0          1         1         1        FALSE
  1         0          1         1         1        TRUE

=== Confusion Matrix ===

  a  b   <-- classified as
 55  0   a = FALSE
  0  6   b = TRUE