What can we learn about the distribution of fitness effects of new mutations from DNA sequence data?

What can we learn about the distribution of fitness effects of new mutations from DNA sequence data?