Spades is a good assembler that unfortunately does not give stats on reads, assembly quality, etc. That said, denovo assembly, unlike mapping reads to a reference takes into account all reads in the data set to generate a consensus. So, in theory, the assembler uses all the reads you supplied during the assembly process. Denovo assembly to my knowledge does not have 'un-aligned reads'.

I asked the same to spades support and got instant response. They also said the same"spades uses all reads larger than the kmer size". I found the number in the assembly log file and it is actually hidden in it.

Hi Rameez,

I just wanted to know why you want see number of reads used in the assembly?

is it really required to assessed quality of denovo assembly?

if yes then u can map raw reads to assembled genome using BWA and SAMTools. this will give you number and percent of reads map on the consensus assembled genome, which eventually give you idea of assembly quality based on the percent of reads mapped. higher the percent of reads mapped higher assembly quality. 

I am wondering you would be more interested in the contig size (Longest contigs, N50) of assembled genome to ensure quality of assembly.

If you have performed prior assembly, quality assessment and filtering and had used only high quality reads for assembly, i guess this is enough.

as far as number of reads used in the assembly, you try other assembly software like SOAPdenovo. 

Hi,

This also happened to me. what i have tried is to map reads back to assembled contigs. But only ~80% of them are aligned.