Seqff is tool for estimating fetal DNA fraction from https://obgyn.onlinelibrary.wiley.com/doi/full/10.1002/pd.4615

• You need to prepare:

  • .bam file of a sample as input for estimating its fetal DNA fraction.
  • All files listed in this repo., and put in the same directory.
  • An R software with Rsamtools, argparse, dplyr, stringr installed.

• And then, just run the following code in the terminal:

$ Rscript seqff.r -f input.bam -o output.txt

Where input.bam is file name of your .bam file, and output.txt is the file which will store the result of seqff.

If your .bam file was aligned to hg38 genome instead of the default hg19 genome, you can run by:

When using hg38 mode, you shall read the disclaimer (on the end of the page) first!

$ Rscript seqff.r -f input.bam -o output.txt --hg38

• You can input .bam files straightly after alignment instead of .sam files without header.
• Sometimes, output of enet would be NA due to NA values in variable bincounts. We set them to zeros before calculating enet value.
• Less arguments needed.
• In some conditions, the chromosome names in .bam file do not contain string 'chr'. This script automatically detects them and adds the 'chr' strings (see Issue #1).
• Hg38 mode have been implemented by setting --hg38 (see help document for details). This mode is achieved by liftOver.

Disclaimer: When using liftOver for conversion, 8.29% of bins cannot correspond to hg38. Therefore, the author cannot guarantee that the hg38 mode necessarily reflects the ideal performance of the model. The author tested a small number of samples (aligned to hg38 genome, using hg38 and hg19 modes respectively) and found that the predicted results remained almost unchanged.