71
Asynchronous Pattern Matching - Metrics Amihood Amir

Asynchronous Pattern Matching - Metrics

Embed Size (px)

DESCRIPTION

Asynchronous Pattern Matching - Metrics. Amihood Amir. Motivation. Motivation. In the “old” days: Pattern and text are given in correct sequential order. It is possible that the content is erroneous. New paradigm: Content is exact, but the order of the pattern symbols may be scrambled. - PowerPoint PPT Presentation

Citation preview

Page 1: Asynchronous  Pattern Matching - Metrics

Asynchronous Pattern Matching -

Metrics

Amihood Amir

Page 2: Asynchronous  Pattern Matching - Metrics

Motivation

Page 3: Asynchronous  Pattern Matching - Metrics

Motivation

In the “old” days: Pattern and text are given in correct sequential order. It is possible that the content is erroneous.

New paradigm: Content is exact, but the order of the pattern symbols may be scrambled.

Why? Transmitted asynchronously? The nature of the application?

Page 4: Asynchronous  Pattern Matching - Metrics

Example: Swaps

Tehse knids of typing mistakes are very common

So when searching for pattern These we are seeking the symbols of the pattern but with an order changed by swaps.

Surprisingly, pattern matching with swaps is easier than pattern matching with mismatches (ACHLP:01)

Page 5: Asynchronous  Pattern Matching - Metrics

Example: Reversals

AAAGGCCCTTTGAGCCC

AAAGAGTTTCCCGGCCC

Given a DNA substring, a piece of it can detach and reverse .

This process still computationally tough.

Question: What is the minimum number of reversals necessary to sort a permutation of 1,…,n

Page 6: Asynchronous  Pattern Matching - Metrics

Global Rearrangements?

Berman & Hannenhalli (1996) called this Global Rearrangement as opposed to Local Rearrangement (edit distance). Showed it is NP-hard.

Our Thesis: This is a special case of errors in the address rather than content .

Page 7: Asynchronous  Pattern Matching - Metrics

Example: Transpositions

AAAGGCCCTTTGAGCCC

AATTTGAGGCCCAGCCC

Given a DNA substring, a piece of it can be transposed to another area .

Question: What is the minimum number of transpositions necessary to sort a permutation

of 1,…,n?

Page 8: Asynchronous  Pattern Matching - Metrics

Complexity?

Bafna & Pevzner (1998), Christie (1998), Hartman (2001): 1.5 Polynomial Approximation.

Not known whether efficiently computable.

This is another special case of errors in the address rather than content .

Page 9: Asynchronous  Pattern Matching - Metrics

Example: Block Interchanges

AAAGGCCCTTTGAGCCC

AAGTTTAGGCCCAGCCC

Given a DNA substring, two non-empty subsequences can be interchanged .

Question: What is the minimum number of block interchanges necessary to sort a

permutation of 1,…,n?

Christie (1996): O(n )2

Page 10: Asynchronous  Pattern Matching - Metrics

SummaryBiology: sorting permutations

Reversals(Berman & Hannenhalli, 1996)

Transpositions(Bafna & Pevzner, 1998)

Pattern Matching:

Swaps(Amir, Lewenstein & Porat, 2002)

NP-hard

?

Block interchanges O(n2)(Christie, 1996)

O(n log m)

Note: A swap is a block interchange simplification1. Block size 2. Only once 3. Adjacent

Page 11: Asynchronous  Pattern Matching - Metrics

Edit operations map

Reversal, Transposition, Block interchange:

1. arbitrary block size 2. not once 3. non adjacent

4. permutation 5. optimization

Interchange:

1. block of size 1 2. not once 3. non adjacent

4. permutation 5. optimization

Generalized-swap:

1. block of size 1 2. once 3. non adjacent

4. repetitions 5. optimization/decision

Swap:

1. block of size 1 2. once 3. adjacent

4. repetitions 5. optimization/decision

Page 12: Asynchronous  Pattern Matching - Metrics

S=abacbF=bbaca

interchange

S=abacbF=bbaac

interchange

matches

S1=bbaca

S2=bbaac

S=abacbF=bcaba

generalized-swap

matches

S1=bbaca

S2=bcaba

Definitions

Page 13: Asynchronous  Pattern Matching - Metrics

Generalized Swap MatchingINPUT: text T[0..n], pattern P[0..m]

OUTPUT: all i s.t. P generalized-swap matches T[i..i+m]Reminder: Convolution

The convolution of the strings t[1..n] and p[1..m] is the string t*p such that:

(t*p)[i]=k=1,m(t[i+k-1]p[m-k+1]) f or all 1 i n-m

Fact: The convolution of n-length text and m-length pattern can be done in O(n log m) time using FFT.

Page 14: Asynchronous  Pattern Matching - Metrics

In Pattern MatchingConvolutions:

O(n log m) using FFT

210

2423222120

1413121110

0403020100

012

43210

rrr

bababababa

bababababa

bababababa

bbb

aaaaab0 b1 b2 b0 b1 b2b0 b1 b2

Page 15: Asynchronous  Pattern Matching - Metrics

Problem: O(n log m) only in algebraically closed fields, e.g. C.

Solution: Reduce problem to (Boolean/integer/real) multiplication. SThis reduction costs!

Example: Hamming distance.

Counting mismatches is equivalent to Counting matches

A B A B C

A B B B A

Page 16: Asynchronous  Pattern Matching - Metrics

Example:

Count all “hits” of 1 in pattern and 1 in text.

011

01001

00000

01001

101

010011 0 11 0 11 0 1

Page 17: Asynchronous  Pattern Matching - Metrics

For a

Define:

)(ba1 if a=b

0 o/w

)()...()()()...( 321321 naaaana SSSSSSSS

Example:

1001100)( abbaabba

Page 18: Asynchronous  Pattern Matching - Metrics

For cba ,,

Do:

)()(

)()(

)()(

Rcc

Rbb

Raa

PT

PT

PT

+

+

Result: The number of times a in pattern matches a in text + the number of times b in pattern matches b in text + the number of times c in pattern matches c in text.

Page 19: Asynchronous  Pattern Matching - Metrics

Idea: assign natural numbers to alphabet symbols, and construct:

T’: replacing the number a by the pair a2,-a

P’: replacing the number b by the pair b, b2.

Convolution of T’ and P’ gives at every location 2i:

j=0..mh(T’[2i+j],P’[j])

where h(a,b)=ab(a-b).

3-degree multivariate polynomial.

Generalized Swap Matching: a Randomized Algorithm…

Page 20: Asynchronous  Pattern Matching - Metrics

Generalized Swap Matching: a Randomized Algorithm…

Since: h(a,a)=0 h(a,b)+h(b,a)=ab(b-a)+ba(a-b)=0,

a generalized-swap match 0 polynomial.

Example:

Text: ABCBAABBC

Pattern: CCAABABBB

1- 1 ,4- 2 ,9- 3,4- 2,1- 1,1- 1,4- 2,4- 2,9- 3 3 9 ,3 9 ,1 1,1 1,2 4 ,1 1,2 4 ,2 4,2 4

3- 9,12 -18,9- 3,4- 2,2- 4,1- 1,8- 8,8- 8,18- 12

Page 21: Asynchronous  Pattern Matching - Metrics

Problem: It is possible that coincidentally the result will be 0 even if no swap match.

Example: for text ace and pattern bdf we get a multivariate degree 3 polynomial:

We have to make sure that the probability for such a possibility is quite small.

0222222 effecddcabba

Generalized Swap Matching: a Randomized Algorithm…

Page 22: Asynchronous  Pattern Matching - Metrics

Generalized Swap Matching: a Randomized Algorithm…

What can we say about the 0’s of the polynomial?

By Schwartz-Zippel Lemma prob. of 0degree/|domain|.

Conclude:

Theorem: There exist an O(n log m) algorithm that reports all generalized-swap matches and reports false matches with prob.1/n.

Page 23: Asynchronous  Pattern Matching - Metrics

Generalized Swap Matching:De-randomization?

Can we detect 0’s thus de-randomize the algorithm?

Suggestion: Take h1,…hk having no common root.

It won’t work,

k would have to be too large !

Page 24: Asynchronous  Pattern Matching - Metrics

Generalized Swap Matching: De-randomization?…

Theorem: (m/log m) polynomial functions are required to guarantee a 0 convolution value is a 0 polynomial.Proof: By a linear reduction from word equality.

Given: m-bit words w1 w2 at processors P1 P2

Construct: T=w1,1,2,…,m P=1,2,…,m,w2.

Now, T generalized-swap matches P iff w1=w2.

Communication Complexity: word equality requires exchanging (m)

bits,

We get: klog m= (m), so k must be (m/log m).

P1 computes:

w1 * (1,2,…,m)

log m bit result

P2 computes:

(1,2,…,m) * w2

Page 25: Asynchronous  Pattern Matching - Metrics

Interchange Distance Problem

INPUT: text T[0..n], pattern P[0..m]

OUTPUT: The minimum number of interchanges s.t. T[i..i+m] interchange matches P.

Reminder: permutation cycle

The cycles (143) 3-cycle, (2) 1-cycle represent 3241.

Fact: The representation of a permutation as a product of disjoint permutation cycles is unique.

Page 26: Asynchronous  Pattern Matching - Metrics

Interchange Distance Problem…

Lemma: Sorting a k-length permutation cycle requires exactly k-1 interchanges.

Proof: By induction on k.

Theorem: The interchange distance of an m-length permutation is m-c(), where c() is the number of permutation cycles in .

Result: An O(nm) algorithm to solve the interchange distance problem.

A connection between sorting by interchanges and generalized-swap matching?

Cases: (1), (2 1), (3 1 2)

Page 27: Asynchronous  Pattern Matching - Metrics

Interchange Generation Distance Problem

INPUT: text T[0..n], pattern P[0..m]

OUTPUT: The minimum number of interchange-generations s.t. T[i..i+m] interchange

matches P.Definition: Let S=S1,S2,…,Sk=F, Sl+1 derived from Sl via interchange Il. An interchange-generation is a subsequence of I1,…,Ik-1 s.t. the interchanges have no index in common.

Note: Interchanges in a generation may occur in parallel.

Page 28: Asynchronous  Pattern Matching - Metrics

Interchange Generation Distance Problem…

Lemma: Let be a cycle of length k>2. It is possible to sort in 2 generations and k-1 interchanges.

Example: (1,2,3,4,5,6,7,8,0)

generation 1:

(1,8),(2,7),(3,6),(4,5)

(8,7,6,5,4,3,2,1,0)

generation 2:

(0,8),(1,7),(2,6),(3,5)

(0,1,2,3,4,5,6,7,8)

Page 29: Asynchronous  Pattern Matching - Metrics

Sorting a General Cycle in Two Rounds

Algorithm:

Exchange contents of locations 2 and n 3 and n-1 4 and n-2

...

Then bring every one to place simultaneously.

Page 30: Asynchronous  Pattern Matching - Metrics

EXAMPLE:Index 6 3 5 1 4 2 Sorted 0 1 2 3 4 5 Permutation 2 5 4 0 1 3

3 4 5 0 1 2

0 4 5 3 1 2 0 1 5 3 4 2 0 1 2 3 4 5

Page 31: Asynchronous  Pattern Matching - Metrics

Why Does it Work?

We want to send 0 to its place

So in the last location should be the number where 0 is

now.

This number is in location 2.

Same reasoning true for 1, 2… ,

Page 32: Asynchronous  Pattern Matching - Metrics

Interchange Generation Distance Problem…

Theorem: Let maxl() be the length of the longest permutation cycle in an m-length permutation . The interchange generation distance of is exactly:

1. 0, if maxl()=1.

2. 1, if maxl()=2.

3. 2, if maxl()>2.

Note: There is a generalized-swap match iff sorting by interchanges is done in 1 generation.

Page 33: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

INPUT: text T[0..n], pattern P[0..m]

OUTPUT: The L1-distance between T[i..i+m] and P for all i[0,n-m].

Where, L1-distance(T[i..i+m],P)=|j-Ti(j)|

Do we need to try all pairings of same letters?

How do we pair the symbols?

Example:

Text: ABCBAABBC

Pattern: CCAABABBB

Page 34: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

n

i

LM iMiTTd

021

1 |)(|),(

Definition: Let T1[0..n], be a string over ∑ and T2[0..n] a permutation of T1.

A pairing between T1 and T2 is a bijection

M:{0,…,n}→{0,…,n}

Where T1[i]=T2[M(i)] , for all i=0,…,n

The L1-distance between T1 and T2 under M is

The L1-distance between T1 and T2 is)(min),( 2,1

121

1 TTdTTd LMM

L

Page 35: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=

Page 36: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8

Page 37: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5=13

Page 38: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5+2=15

Page 39: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5+2+4=19

Page 40: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5+2+4+1=20

Page 41: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5+2+4+1+3=23

Page 42: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5+2+4+1+3+5=28

Page 43: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5+2+4+1+3+5+7=35

Page 44: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M1

T2= C A C B B B B A A

dM1L1(T1,T2)=8+5+2+4+1+3+5+7+5=40

Page 45: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=

Page 46: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1

Page 47: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2=3

Page 48: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2+2=5

Page 49: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2+2+4=9

Page 50: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2+2+4+1=10

Page 51: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2+2+4+1+5=15

Page 52: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2+2+4+1+5+2=17

Page 53: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2+2+4+1+5+2+5=22

Page 54: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem

EXAMPLE:

T1= A B B A B C A C B

M2

T2= C A C B B B B A A

dM2L1(T1,T2)=1+2+2+4+1+5+2+5+2=24

Note that dM2L1(T1,T2)=40

How do we choose the pairing function?

Page 55: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem…Fortunately, we know how an optimal pairing looks...

lemma: Let T1,T2∑m be two strings s.t. T2 is a permutation of T1. Let M be the pairing function that, for all a and k, moves the k-th a in T1 to the location of the k-th a in T2. Then,

dL1(T1,T2)=dML1(T1,T2)

Result: An O(nm) algorithm to solve the L1-distance problem.

Example: An optimal pairing

Text: ABCBAABBC

Pattern: CCAABABBB

Page 56: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem…Proof of pairing lemma:

Note that in M, for a fixed alphabet symbol a, there are no “crossovers”.

We will show that crossovers do not help.

Cases: M M’

1)

x y z

Cost in M=x+y+y+z

Cost in M’=x+y+z+y

Page 57: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem…Proof of pairing lemma (continued…):

2) M M’

x y z

Cost in M=x+z

Cost in M’=x+y+y+z

Page 58: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem…Proof of pairing lemma (continued…):

3) M M’

x y z

Cost in M=x+z

Cost in M’=x+y+z+y

Page 59: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem…For pattern with distinct entries we can do better...

Idea: Use computations from position i to position i+1.Example:

Text: 2312331

Pattern: 123

Dist(1)=|1-3|+|2-1|+|3-2|=4

Direction relative to pattern:

Left Match Right

2,3 1

Page 60: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem…

Dist(2)=?

Now we move to the next location…

Text: 2312331

Pattern: 123

Direction relative to pattern:

Left (+1 to Dist) Match Right (-1 to Dist)

2,3 1,2

Page 61: Asynchronous  Pattern Matching - Metrics

L1-Distance Problem…

Dist(2)=Dist(1)+|Left|-|Right|-left most+right most=

=4+1-1-1+|3-2|=4 (=|1-2|+|2-3|+|3-1|)

Result: An O(n) algorithm to solve the L1-distance problem for pattern with distinct entries.

Next location distance computation…

Text: 2312331

Pattern: 123

Page 62: Asynchronous  Pattern Matching - Metrics

L2-Distance Problem

n

i

LM iMiTTd

0

221

2 ))((),(

Definition: Let T1[0..n], be a string over ∑ and T2[0..n] a permutation of T1.

A pairing between T1 and T2 is a bijection

M:{0,…,n}→{0,…,n}

Where T1[i]=T2[M(i)] , for all i=0,…,n

The L2-distance between T1 and T2 under M is

The L2-distance between T1 and T2 is)(min),( 2,1

221

2 TTdTTd LMM

L

Page 63: Asynchronous  Pattern Matching - Metrics

L2-Distance ProblemINPUT: text T[0..n], pattern P[0..m]

OUTPUT: The L2-distance between T[i..i+m] and P for all i[0,n-m].

Where, L2-distance(T(i),P)=|j-MT(i)(j)|2

Do we need to try all pairings of same letters?No, the pairing lemma works here too.

Result: An O(nm) algorithm to solve the L2-distance problem.

We can do better…

Page 64: Asynchronous  Pattern Matching - Metrics

L2-Distance Problem…

Idea: Consider T and P of same length.

Let lista(P) and lista(T) for each letter a, be the list of locations in which a occurs in P and T.

The L2-distance between T and P is:

dL2(T,P)=aPj[0,…|list(a)|](lista(P)[j]-lista(T(i))[j])2

Schematically: Fix letter a:

T

P

list(T) = i1, i2, …, ik

list(P) = j1, j2, …, jk

k

lll

L jiPTd1

22 )(),(

Page 65: Asynchronous  Pattern Matching - Metrics

L2-Distance Problem…

k

lll

L jiPTd1

22 )(),(

k

lllll jjii

1

22 )2(

k

l

k

l

k

lllll jiji

1 1 1

22 2

easyto

compute

easyto

compute

convolution

Page 66: Asynchronous  Pattern Matching - Metrics

L2-Distance Problem…large text

Idea: Consider lista(P) and lista(T(i)) for each letter a, the list of locations in which a occurs in P and T(i).

The L2-distance between T(i) and P is:

dL2(T(i),P)=aPj[0,…|list(a)|](lista(P)[j]-lista(T(i))[j])2

Now, we want to use lista(T) instead of lista(T(i)).

Now the text location indices are not fixed…

Page 67: Asynchronous  Pattern Matching - Metrics

L2-Distance Problem…large text

Schematically: Fix letter a:

T x

P

list(T) = i1, i2, …, ik,...,ir

list(P) = j1, j2, …, jk

For location x, the text indices need to be i2-x, i3-x, …

Convolution formula:

k

lll

xL jxiPTd1

2)(2 ))((),(

Page 68: Asynchronous  Pattern Matching - Metrics

L2-Distance Problem…large text

k

lll

xL jxiPTd1

2)(2 ))((),(

k

lll

k

ll

k

ll jxijxi

11

2

1

2 )(2)(

k

llll

k

l

k

l

k

llll xjjijixkxi

11 1 1

222 )(22

k

l

k

llll

k

l

k

l

k

llll jxjijixkxi

1 11 1 1

222 222

easy to compute convolutioneasy tocompute

Page 69: Asynchronous  Pattern Matching - Metrics

L2-Distance Problem…large alphabet

)log( mnO i

Problem: What happens when unbounded number of alphabet symbols? How many convolutions?

Let ∑={a1,…,ak} and let ni be the number of times ai occurs in T, i=1,…,k.

Clearly, n1+n2+…+nk=n.

Time: For each symbol ai:

Total Time:

Result: An O(n log m) algorithm to solve the L2-distance problem.

)log()log)(()log(1 1

mnOmnOmnOk

l

k

lii

Page 70: Asynchronous  Pattern Matching - Metrics

Open Problems

1. Interchange distance faster than O(nm)?

2. Asynchronous communication – different errors in address bits.

3. Different error measures than interchange/block interchange/transposition/reversals for errors arising from address bit errors.

Note: The techniques employed in asynchronous pattern matching have so far proven new and different from traditional pattern matching.

Page 71: Asynchronous  Pattern Matching - Metrics

The End