Families with multiple instances of lymphoid cancers are rare. In suchfamilies, members generally develop different types of lymphoid cancer, suchas non-Hodgkin lymphoma, Hodgkin lymphoma, chronic lymphocytic leukemia andmultiple myeloma; rather than developing the same disease. The presence ofdifferent lymphoid cancers in the same family implies that there are sharedsusceptibility factors for these cancers, including shared geneticfactors. In these families, which show phenotypic heterogeneity andincomplete penetrance, we anticipate a need to analyze data both within andbetween families to identify cancer susceptibility genes. Towards this goal,we consider various ways to define the probability that genetic variants areshared in particular subsets of affected family members, expanding on methodsdeveloped by Bureau et al. (2014). These methods will allow prioritizationof rare variants identified by exome sequencing in over 100 lymphoid cancerfamilies. Our aim is to model rare variants observed in the affectedindividuals to allow prioritization for validation and functional studies. Akey feature of our intended model is a weight that prioritizes variantsshared by subsets of the affected members within each family. We propose andexplore some candidates for this weight as they apply to a particularlyinteresting lymphoid cancer family.