<s>
DIALIGN-TX	B-Application
is	O
a	O
multiple	O
sequence	O
alignment	B-Application
program	I-Application
written	O
by	O
Amarendran	O
R	O
.	O
Subramanian	O
and	O
is	O
substantial	O
improvement	O
of	O
DIALIGN-T	B-Application
by	O
combining	O
greedy	O
and	O
progressive	O
alignment	O
strategies	O
in	O
a	O
new	O
algorithm	O
.	O
</s>
<s>
The	O
original	O
DIALIGN-T	B-Application
is	O
a	O
reimplementation	O
of	O
the	O
multiple-alignment	O
program	O
DIALIGN	O
.	O
</s>
<s>
However	O
,	O
like	O
the	O
original	O
implementation	O
of	O
the	O
program	O
,	O
DIALIGN-T	B-Application
uses	O
a	O
straightforward	O
greedy	O
approach	O
to	O
assemble	O
multiple	O
alignments	O
from	O
local	O
pairwise	O
sequence	O
similarities	O
.	O
</s>
<s>
DIALIGN-TX	B-Application
is	O
a	O
substantial	O
improvement	O
of	O
DIALIGN-T	B-Application
that	O
combines	O
the	O
previous	O
greedy	O
algorithm	O
with	O
a	O
progressive	O
alignment	O
approach	O
.	O
</s>
