ichorCNA can be run without any reference samples and a panel of normals is not necessary for analysis with ichorCNA. However, if you choose, you can use a normal reference or a panel of normals.

We provide a panel of normals (PoN) with ichorCNA but generating your own using samples that were processed and sequenced similarly to your cancer patient cfDNA samples may reduce noise and improve accuracy. These data help to further normalize the cancer patient cfDNA to correct for systematic biases arising from library construction, sequencing platform, and cfDNA-specific artifacts. We also provide an R script to generate a PoN with your own cfDNA lowpass samples.

Create a WIG file for each sample in your PoN just as you would for any cfDNA sample you would analyze with ichorCNA.

/path/to/HMMcopy/bin/readCounter --window 1000000 --quality 20 \
    --chromosome "1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,X,Y" \
    /path/to/tumor.bam > /path/to/tumor.wig

Use the createPanelOfNormals.R script provided in the scripts directory to generate your PoN. As input, this script takes a file that has the path to each WIG file you'd like to use in your panel (one per line, no header).

Rscript createPanelOfNormals.R 
    --filelist /path/to/wig_files.txt \
    --gcWig /path/to/gc.wig --mapWig /path/to/map.wig \
    --centromere /path/to/centromeres_file.txt \
    --outfile base_outfile_name

• --filelist /path/to/wig_files.txt - file containing a list of the paths to all the normals in the panel to analyze
• --gcWig /path/to/gc.wig - GC Wig file for reference genome (e.g. ichorCNA/inst/extdata/gc_hg38_1000kb.wig)
• --mapWig /path/to/map.wig - Mappabiliy Wig file for reference genome (e.g. ichorCNA/inst/extdata/map_hg38_1000kb.wig)
• --centromere /path/to/centromeres_file.txt - File containing Centromere locations (e.g. GRCh38.GCA_000001405.2_centromere_acen.txt)
• --exons.bed - Can be used if uses which to generate a panel for targeted sequences such as exome capture.

When you run ichorCNA, you can pass the .rds output file in using the --normalPanel option to normalize your sample using the PoN.