Microsatellite instability characterization

1. A method comprising:

determining a microsatellite stability background model based on a set of observed variant allele frequencies from a first set of samples associated with microsatellite stability, wherein determining the microsatellite stability background model comprises performing a scale normalization based on the observed variant allele frequencies;

determining a set of loci associated with microsatellite instability based on the microsatellite stability background model and a second set of samples associated with the microsatellite instability;

determining a microsatellite instability threshold based on the set of loci and a third set of samples associated with the microsatellite stability and the microsatellite instability;

determining a microsatellite instability parameter for a user associated with a user sample, based on the user sample and the set of loci;

determining a microsatellite instability characterization of high (MSI-H), associated with the at least one cancer condition, for the user based on a comparison between the microsatellite instability parameter and the microsatellite instability threshold; and

administering at least one of pembrolizumab, durvalumab, avelumab, atezolizumab, or nivolumab to the user having a determined microsatellite instability characterization of MSI-H.

2. The method of claim 1 , wherein determining the microsatellite instability parameter comprises:

determining a first distance based on the microsatellite stability background model and the user sample;

determining a second distance based on a microsatellite instability background model and the user sample; and

determining the microsatellite instability parameter based on the first and the second distances.

3. The method of claim 2 , wherein determining the microsatellite instability parameter comprises:

determining a set of microsatellite stability distances comprising the first distance, based on the microsatellite stability background model and the user sample, wherein each microsatellite stability distance of the set of microsatellite stability distances corresponds to a different locus of the set of loci;

determining a set of microsatellite instability distances comprising the second distance, based on the microsatellite instability background model and the user sample, wherein each microsatellite instability distance of the set of microsatellite instability distances corresponds to a different locus of the set of loci; and

determining the microsatellite instability parameter based on the set of microsatellite stability distances and the set of microsatellite instability distances.

4. The method of claim 3 , wherein the third set of samples is associated with a plurality of cancer types comprising the at least one cancer condition.

5. The method of claim 3 , wherein the set of microsatellite stability distances and the set of microsatellite instability distances correspond to a Mahalanobis distance type.

6. The method of claim 5 , wherein determining the microsatellite instability parameter comprises:

determining a per-locus log-likelihood ratio for each locus of the set of loci based on the set of microsatellite stability distances and the set of microsatellite instability distances corresponding to the Mahalanobis distance type; and

determining the microsatellite instability parameter based on the per-locus log-likelihood ratios.

7. The method of claim 5 , wherein determining the set of loci comprises:

determining a set of Mahalanobis distances corresponding to the Mahalanobis distance type, based on the microsatellite stability background model and the second set of samples associated with the microsatellite instability; and

determining the set of loci based on the set of Mahalanobis distances.

8. The method of claim 2 , wherein determining the microsatellite instability characterization for the user as MSI-H comprises determining a binary status determination for the microsatellite instability based on the comparison between the microsatellite instability parameter and the microsatellite instability threshold.

9. The method of claim 1 , wherein performing the scale normalization comprises normalizing observed variant allele frequencies of the set of observed variant allele frequencies based on a single-base deletion variant allele frequency.

10. The method of claim 1 , wherein determining the microsatellite stability background model comprises determining the microsatellite stability background model based on the set of observed variant allele frequencies determined from sequencing nucleic acid material from the first set of samples using a next-generation sequencing system associated with homopolymer-associated errors.

11. A method comprising:

determining a microsatellite instability parameter for a user associated with a user sample, based on the user sample and a set of loci associated with microsatellite instability, wherein determining the microsatellite instability parameter comprises:

determining a microsatellite stability distance for the user based on the user sample and a microsatellite stability background model associated with first variant allele frequencies from first samples associated with microsatellite stability, and

determining a microsatellite instability distance for the user based on the user sample and a microsatellite instability background model associated with second variant allele frequencies from second samples associated with the microsatellite instability;

determining a microsatellite instability characterization of high (MSI-H) for the user based on the microsatellite instability parameter; and

administering at least one of pembrolizumab, durvalumab, avelumab, atezolizumab, or nivolumab to the user having a determined microsatellite instability characterization of MSI-H.

12. The method of claim 11 , wherein the set of loci is determined based on the microsatellite stability background model and a set of samples associated with the microsatellite instability.

13. The method of claim 12 , wherein the set of loci is determined based on a set of Mahalanobis distances derived from the microsatellite background model and the set of samples associated with the microsatellite instability.

14. The method of claim 12 , wherein the set of loci is determined based on the microsatellite stability background model, the set of samples associated with the microsatellite instability, and at least one of average read depth, tract length, and typical strand bias.

15. The method of claim 12 , wherein the set of samples is associated with a plurality of cancer types.

16. The method of claim 1 , wherein the microsatellite stability background model is determined based on a normalization of observed variant allele frequencies with a single-base deletion variant allele frequency.

17. The method of claim 11 , wherein determining the microsatellite instability characterization as MSI-H comprises classifying the user with the microsatellite instability based on the microsatellite instability parameter and a microsatellite instability threshold associated with the set of loci.

18. The method of claim 11 , wherein determining the microsatellite instability parameter as MSI-H comprises determining the microsatellite instability parameter based on the set of loci and sequencing of the user sample using a semiconductor-based sequencing system associated with homopolymer-associated errors.