Alligning to assembly without chromosomes

Question

[jesse.hoff]

I'm working on aligning some data to an assembly that has been released without final placement of contigs into scaffolds and scaffolds into chromosomes. that data may be forthcoming but in the meantime the reference is a bunch of contigs. Still, I have some illumina paired end dna 100bp that I'm aligning using bwa mem. currently about 4% of mate pairs are mapping to different contigs. Our goal is variant calling for analysis of selection signatures, among other things

Any thoughts about bwa and other downstream parameters that optimize for the contig structure of the data?

Thanks