I'm not sure if recombination rates have anything to do with it. You've sequenced a locus with two reads (F/R) and want to assemble those reads. Whatever level of recombination, those reads are true data. Just assemble with normal parameters.
phred *.ab1 -sd . -qd . -pd . phd2fasta forward.phd.1 reverse.phd.1 -os locus.fna -oq locus.fna.qual phrap locus.fna -ace 2>phrap.err 1>phrap.out # output is in locus.fna.contigs and locus.fna.contigs.qual
If these forward and reverse sequences do not assemble, consider the possibility that there is some physical obstruction to their overlap. Maybe a recombination event created an insertion between the forward and reverse sequences, such that you need to sequence the insert before you can assemble all sequences. Or, your forward or reverse primer locus was recombined to a different place in the genome. To overcome this obstruction, you may need to primer walk.