The simplest way is to transpose the data, split out the last letter from the q variables, then re-transpose.
data have;
input id q1a q2a q1b q2b q1c q2c;
datalines;
1 3 0 1 1 1 9
2 4 9 1 2 2 0
3 5 9 1 2 4 0
;
run;
proc transpose data=have out=temp1;
by id;
run;
data temp2;
set temp1;
length type $1;
type=substr(_NAME_,3);
_NAME_=substr(_NAME_,1,2);
run;
proc transpose data=temp2 out=want (drop=_:) ;
by id type;
id _NAME_;
var COL1;
run;