I tried increase the memory for SPADES by specifying

withName: SPADES {
        cpus   = { 48 }
        memory = { 440.GB }
        time   = { 4.d }
    }

But the job was pending for a while, then I found out (as below) it was because it reached the CPU run limit if I understood correctly, so I changed it back and still couldn't proceed the pipeline.

JOBID PARTITION     NAME     USER ST       TIME  NODES NODELIST(REASON)
          11587200    memory nf-NFCOR junhongj PD       0:00      1 (AssocGrpCPURunMinutesLimit)