UPDB - Analytical Services

UPDB offers comprehensive analytical consulting services to assist investigators with analysis of UPDB data. These services are provided by our team of professional biostatisticians and data scientists who have deep knowledge of the database and extensive experience with the unique software applications designed for analysis of UPDB data.
Services include experimental design, survival analysis, geospatial analysis, genetic data analysis, familial relative risk analysis, familial standardized incidence ratio analysis, pedigree visualization and other applications. A few of the most common applications are described below.

Familial Relative Risk

Familial Relative Risk (FRR) analysis is a method to assess whether a phenotype is more or less likely to occur among the relatives of affected individuals versus the relatives of matched controls. The analysis results may help to determine if there is an inherited familial risk for the phenotype or disease being studied.

FRR results are reported as a relative risk estimate and an associated p-value for each category of relatives that is studied. FRR is calculated separately for first-degree relatives, second-degree relatives, third-degree relatives, or other specific categories. The calculations can be performed with either logistic regression or Cox Proportional Hazards models.

Familial Standardized Incidence Ratio (FSIR)

The Familial Standardized Incidence Ratio (FSIR) is a method to identify individuals whose family members or descendants have a greater incidence of a disease or other phenotype than would be expected based on observed population incidence rates. This method is made possible by UPDB’s powerful combination of longitudinal population health and pedigree data. FSIR analysis can help to identify families and individuals that may have increased genetic risk for a disease.

The FSIR method is similar to the standardized incidence ratio (SIR) method commonly used in cancer epidemiology and other fields. FSIR is a special case of SIR, where the study group is defined as the known relatives or descendants of a particular person:

FSIR = (observed disease incidence in a family) / (expected disease incidence in the family based on population rate).

FSIR analysis results include the FSIR value, an associated p-value, statistics about pedigree size, observed and expected case counts, and other data. FSIR results are usually calculated for pedigree founders and based on the disease incidence among their direct descendants. This approach is useful to identify pedigrees that have an excessive number of disease cases among the descendants of a common ancestor, potentially due to inherited genetic risk. FSIRs can also be calculated for individual probands, considering disease incidence among all known biological relatives, as a method to quantify familial disease risk at an individual person level.

Survival Analysis

Survival analysis is used to associate variables with mortality or other time to event outcomes. Cox proportional hazards regression is the most common method that we use and results in hazard ratios (with p-values) to describe the changes in the rate of an event depending on the input variable. Examples of time to event outcomes include time from diagnosis to death or censoring and time from diagnosis to first treatment. Examples of input variables are various diagnoses, environmental exposure, and interventions.