Counting a value within a specific range in array in SAS - arrays

Question

So my dataset looks like this:
ABC1 ABC2 ABC3 ABC4 ABC5 DEF1 DEF2 DEF3 DEF4 DEF5
1 0 0 1 . 0 1 1 0 .
I want my output to be:
XYZ1 XYZ2 XYZ3 XYZ4 XYZ5
0 1 1 0 .
Basically if DEF2 = 1 and count of ABC3 and ABC4 and ABC5 of 1 is > 0 then XYZ2 is 1.
I have tried the following code but it doesnt work
data want;
set have;
array ABC ABC:;
array DEF DEF:;
array XYZ [5] $1;
do i = 1 to dim(ABC)-5;
if ABC(i) = . then XYZ(i) = '';
else if (DEF(i) = 1 and sum(ABC(i+1), ABC(i+3)) > 0) then XYZ(i) = 1;
else XYZ(i) = 0;
end;
drop i;
run;

Answer 1

Lets pivot things for a better understanding
index ABC DEF XYZ (wanted)
----- --- --- ---
1 1 0 0 (because DEF=0)
2 0 1 1 (sum ABC index 2..5 because DEF=1 # index 2)
3 0 1 1 (sum ABC index 3..5 because DEF=1 # index 3)
4 1 0 0 (because DEF=0)
5 . . . (because DEF=.)
Now apply that understanding to processing variables of the row when arrayed. The items will be processed from 5 to 1, so a running_sum can be computed and applied when necessary.
data want;
set have;
array abc abc:;
array def def:;
array xyz(5);
running_sum = .;
do index = dim(abc) to 1 by -1;
if not missing(abc(index)) then running_sum + abc(index);
if def(index) in (., 0)
then xyz(index) = def(index);
else xyz(index) = running_sum;
end;
run;
Not all processing rules are stated in the question, such as
the case abc(j) = . and abc(k) ne . and k > j
Such a case may never happen.