Objective

1
Solving the Longest Common Subsequence Problem
Using the Associative ASC Processor and
Reconfigurable 2D-Mesh Virdi, Sabegh Singh, Hong
Wang and Robert A. Walker Department of Computer
Science, Kent State University, Kent, OH 44240


LCS Algorithm on reconfigurable 2D-Mesh
LCS Algorithm on reconfigurable 2D-Mesh
Introduction DNA matching is a computationally
demanding task, as the emerging field of
bioinformatics continues to expand the ability
to rapidly search large databases of genetic
information is becoming increasingly important.
One simple form of DNA sequence comparison can be
done by solving the Longest Common Subsequence
(LCS) problem.

• Initially all the switch of all PEs are kept
  open
• Each PEs have a Match Register M and Length
  register L,
• all set to Zero
• Transmit the Text String to First row,
  Replicates the same in all
• rows
• Transmit the pattern String to First Column,
  Replicate the same
• in all columns
• Each PE compares the value held in his
  internal Register, and set
• to 1 if they are equal, else set it to Zero

Objective To design an efficient parallel
algorithm for calculating the longest common
subsequence between two strings. We examine two
cases one for exact match and another for
approximate match



Snapshot of LCS algorithm for Approximate match
Coterie Network

• Developed by Weems and others in 90s
• Used extensively for Image processing
  application, along with others
• PEs form Coterie based upon some condition,
  dictated during execution of algorithm
• Initially had 4 ports (N, S, E, and W)
• It is assumed information propagates at
  electrical speed

• Results
• Constant time algorithm for exact match. An
  Altera APEX1000KC FPGA was sufficiently large
  enough to hold the 6 x11 array of PEs and ran at
  a clock speed of 37 MHz.

A 5 x 5 Coterie Network
Reconfigurable Network in the ASC Processor

• Scalable design with Reconfigurable
• network
• Key to reconfigurability is the Data
• Switch inside each PE
• Data switch has bypass mode to
• allow PE communication to skip
• non-responder, so as to support
• Associative computing

• Conclusions
• Scalable design to handle larger string
• Uses Inexpensive Hardware
• Can be reconfigured for new techniques in future
• Future work will concentrate on providing
  additional parameters to find the best common
  subsequence instead of simply the longest
  common subsequence. In molecular biology, the
  sequence comparison may need to be parameterized
  to give different weights to different nucleotide
  matches or to consider the length or number of
  the individual subsequences and gaps.

Snapshot of LCS algorithm for exact match