Update README and -f

README.md

 A C program, ngscheckmate_fastq, can be directly called to generate a VAF file from one FASTQ file (single-end sequencing) or two FASTQ files(paired-end sequencing). 
 
 Then, another script, vaf_ncm.py is used to read a set of VAF files to complete the downstream analysis. When you need to analyze many FASTQ files, the first VAF file generation using ngscheckmate_fastq can be parallelized.
+
+**If you want to analyze the correlation of samples over multiple runs, I suggest you to save the historical `vaf` files and download the NGSCheckMate from https://github.com/parklab/NGSCheckMate and then run vaf_ncm.py locally.**
 ### Getting Started
 
 We recommend using choppy system and Aliyun OSS service. The command will look like this:
 }
 ```
 
-#### subsampling rate
-The default subsampling rate is 0.3.
+#### other parameters
+subsampling_rate: The default subsampling rate is 1. The speed is not very slow.
+-f in vaf_ncm.wdl: Use strict VAF correlation cutoffs. Recommended when your data may include related individuals (parents-child, siblings).

defaults

 {
-  "subsampling_rate": "0.3",
+  "subsampling_rate": "1",
   "cluster_config": "OnDemand bcs.a2.3xlarge img-ubuntu-vpc",
   "disk_size": "100",
   "docker": "registry.cn-shanghai.aliyuncs.com/pgx-docker-registry/ngscheckmate:v1.0.0"

tasks/vaf_ncm.wdl

     mkdir -p /ncm/fastq_vaf/
     cp ${sep=" " vaf} /ncm/fastq_vaf/
 
-    python /opt/NGSCheckMate/vaf_ncm.py -I /ncm/fastq_vaf/ -O '.'
+    python /opt/NGSCheckMate/vaf_ncm.py -I /ncm/fastq_vaf/ -O '.' -f
   >>>
 
   runtime {