## Set the input separator to be '>'; and the output separator to be "\n"'
BEGIN { $/ = ">"; $\ = "\n"; }

## read each input record (terminated by the '>' character) into $_
LINE: while (defined($_ = <ARGV>)) {

## Remove the record terminator ('>')
    chomp $_;

## split the record on whitespace into @F
    our(@F) = split(' ', $_, 0);

## If the record was empty (the first one always will be with your data),
## then skip to the next record.
    next unless @F;

## remove the 'length=' prefix from the second field
    s/length=// foreach ($F[1]);

## replace the 6th field in the array, with the concatenation of itself 
## and all the fields that follow it
    @F[5] = join('', @F[5..99]);

## Then discard all the fields that follow it.
## Ie. truncate the array after the 6th field
    $#F = 5;

## And print out the fields joined with ','s and prefixed with '>' 
## (and ending with a newline. See output separator above.)
    print '>', join(',', @F);
}