MendelProb Manual

MendelProb is an R package to determine the probability of observing a minimum number of potentially pathogenic variants within a gene in multiple families and/or cases and to also establish how many probands need to be screened to detect multiple observations of potentially pathogenic variants within a gene.

MendelProb can be used for the design of Mendelian disease sequencing studies (candidate gene, exome and whole genome) and grant proposals. For genetic studies, it is necessary to perform power calculations. Although for Mendelian diseases the power of detecting linkage for pedigree(s) can be determined, it is also of great interest to determine the probability of identifying multiple pedigrees or unrelated cases with potentially pathogenic variants in the same gene or estimate the number of probands which need to be screened to detect potentially pathogenic variants. For many Mendelian diseases, due to extreme locus heterogeneity this probability can be small or the number of probands which need to be screened very large. If only one family is observed segregating a variant classified as potentially pathogenic or of unknown significance, the gene cannot be implicated in disease etiology. The probability of identifying additional Mendelian disease families or cases is dependent on the prevalence of the disease due to a specific gene and the sample size to be screened. The observation of additional disease families or cases with potentially pathogenic variants in the same gene as well as evidence of pathogenicity from other sources, e.g., expression and functional studies, can aid in implicating a gene in disease etiology.

Installation

install.packages("devtools")

If you are asked to select a CRAN mirror, choose a mirror site near your current location.

library("devtools")
install_github('statgenetics/mendelprob')
library("mendelprob")

Probability Calculations

The mendel_prob function can be used to determine the probability of detecting a minimum number of probands with potentially pathogenic variants in the same gene. The proband can be either an affected family member or a case. Even if sequence data is generated on more than one family member to perform filtering, when calculating probabilities or number of subjects to be sequenced, each family is counted only once and neither size nor structure of the family will impact the results.

MendelProb can also be used to determine the probability of detecting a minimum number of potentially pathogenic variants within a gene for different data types where it is required that a minimum number of potentially pathogenic variants are observed for one data type e.g., identifying at least three potentially pathogenic variants within a gene where at least one of the variants is observed in a proband who is a member of a family which can establish linkage.

Examples:

mendel_prob(num_probands=625, 
            gene_freq=0.005,
            min_num_variants=2)

mendel_prob(num_probands=125, 
            num_probands_type_2=500, 
            gene_freq=0.005,
            min_num_variants=2, 
            min_num_probands_variants=1)

Where

• num_probands
  • is the total number of probands or if probands are being drawn for two types of data, e.g. families and unrelated cases, this is the number of probands of type I. For this example, there are 125 probands. If for each family two affected individuals were sequenced and there are 125 families, the num_probands would still be 125 since even if there are multiple affected individuals sequenced for a pedigree each family is only counted once.
• num_probands_type_2
  • the number of probands of type 2.
• gene_freq
  • the percent of disease explained by a gene.
• min_num_variants
  • the minimum number of probands with potentially pathogenic variants in the same gene which need to be observed in the total sample.
• min_num_probands_variants
  • the minimum number of probands of the first type (in this example 125) which are required to be observed with potentially pathogenic variants in the same gene.
• note: min_num_probands_variants cannot be greater than min_num_variants.

Sample Size Calculations

The mendel_sample_size functio can be used to determine the number of probands which need to be screened to detect a minimum number of potentially pathogenic variants within the same gene for a specified probability. It can also be used to determine the number of probands which need to be screened if it desired that a minimum number of probands of a certain type are observed with potentially pathogenic variants. For this situation, the proportion of probands of each type which are planned to be sequenced is provided, e.g., one-third of the probands will be from Mendelian families with multiple affected members and two-thirds of the probands are individual cases without additional affected family members.

Examples:

mendel_sample_size(prob=0.8, 
                   gene_freq=0.01,
                   min_num_variants=2)

mendel_sample_size(prob=0.8, 
                   proband_prop=0.3,
                   gene_freq=0.005,
                   min_num_variants=3, 
                   min_num_probands_variants=1) 

Where

• prob
  • the probability of detecting the desired minimum number of potentially pathogenic variants.
• proband_prop
  • the proportion of probands of type I.
• gene_freq
  • the percent of disease explained by a gene.
• min_num_variants
  • the minimum number of probands with potentially pathogenic variants in the same gene which need to be observed for the entire sample.
• min_num_probands_variants
  • the minimum number of potentially pathogenic variants which need to be observed in probands of type I (in this example 1).
• note: min_num_probands_variants cannot be greater than min_num_variants.
• note: The probability and sample size estimations are performed assuming there is no experimental error, i.e. a variant which is not identified or a genotype which is incorrectly called so that no alternative allele is identified – a false negative call. False negative calls will cause probabilities to be overestimated and sample sizes underestimated and will vary depending on the platform which is used to generate sequence data. In most cases the false negative variant call rate is negligible.

Reference & Contact Information

• Please refer to ?mendel_prob or ?mendel_sample_size in R console for detailed function descriptions.
• For comments and bug reports, feel free to create an issue on Github.
• If you found the tool is useful, please cite XXX .
• For any questions and suggestions, feel free to contact
  • Suzanne Leal, [email protected]
  • Zongxiao He, [email protected]