Inference of complex landscapes

In the previous section, we illustrate of how to visualize a complete combinatorial landscape when it is perfectly known. However, we usually only have access to noisy or indirect measurements of such genotype-genotype maps. Moreover, not all possible sequences are evaluated, especially when dealing with high-throughput experiments. In this section, the user will learn how to use our methods to infer [1] complete combinatorial landscapes from both experimental data and observations of natural sequences [2] using our exact Gaussian Process regression techniques.

• Experimental data
• Natural sequences

References

[1]

Zhou J, Wong MS, Chen WC, Krainer AR, Kinney JB, McCandlsih DM. Higher order epistasis and phenotypic prediction (2022)

[2]

Chen WC, Zhou J, Sheltzer JM, Kinney JB, McCandlish DM. Field theoretic density estimation for biological sequence space with applications to 5’ splice site diversity and aneuploidy in cancer (2021)