## Problem 168 — rank:237

**Title:** Distribution of fitness effects

**Added binding assumptions (LaTeX):**

```latex
Permit zero mutant fitness explicitly when lethal mutations are included: the reference $W(g,e)$ is positive, but $W(g+m,e)$ may be zero, producing an atom at $s=-1$. The sampled mutation kernel and ascertainment mechanism determine the DFE, so the target is not the frequency distribution among surviving observed mutants. Uniform CDF accuracy in a withheld context requires declared transfer restrictions and a width/sample-size guarantee, not merely a hierarchical fit.
```

**Direct edits:** 1. **Fresh external checks:** None; supplied sources retained as provenance.