Apparent difference between active region algorithm in mutect.pdf and Mutect2Engine.java:492

Hello,

First off, thank you so much for an excellent toolkit and brilliant forum.
Both have and continue to help me out so much in my work. I am very grateful.

My question relates to an apparent difference in MuTect, GATK4, between the algorithm for identifying active regions (as outlined in docs/mutect/mutect.pdf, page 2) and how it is implemented in the MuTect source file tools/walkers/mutect/Mutect2Engine.java, line 492.

In the code (Mutect2Engine.java:492), fTildeRatio is evaluated as,

```
fTildeRatio = FastMath.exp(MathUtils.digamma(nRef + 1) - MathUtils.digamma(nAlt + 1));
```

However, in mutect.pdf, page 2, just after equation 1, it appears to indicate it should be evaluated as,

```
fTildeRatio = FastMath.exp(MathUtils.digamma(nRef + 1) - MathUtils.digamma(n + 2));
```

Is there a dependency there and if so how and in what situations would it affect the ultimate TLOD value in the vcf file?

Apologies in advance if I'm missing something here or it has been answered before.
I haven't come across it as yet.

Thanks again and best regards, Brian