Computational protocol: Expanding the toolbox of ADHD genetics. How can we make sense of parent of origin effects in ADHD and related behavioral phenotypes?

Similar protocols

Protocol publication

[…] One aspect of POE, the examination of which has been attempted the most, is imprinting, or non-equivalence, of parental alleles. To date, all studies of POE in ADHD have applied transmission disequilibrium test (TDT) based approaches to identify biased transmission of parental alleles [, ]. In short, a 2 × 2 table containing transmitted/non-transmitted allele counts according to the maternal or paternal origin of the alleles is constructed and Fisher’s exact test or Chi square test is applied to compare those. A similar approach is also implemented in the widely used PLINK’s parent-of-origin test []. Although the TDT has been proven to be a useful tool in POE analyses [], this method has some weaknesses: (1) null hypothesis examined the hypothesis tested by TDT itself is that of random meiotic selection of a parental allele for transmission to an offspring, but in POE it is the equality of paternal and maternal transmissions; (2) non-independence of observations the TDT statistics assumes that observed transmissions are independent (under Mendel’s second law), meaning the TDT-based parental transmission comparison may not be valid because of the statistical dependence between maternal and paternal transmissions in triads with two heterozygous parents or because of the possible prenatal maternal effect occurring together with genetic association in triads with homozygote parents [, ]; (3) bias due to missing data POE detection by TDT-based approach may introduce likely bias when incomplete trios are utilized (for example, a trio with genotyped mother only will contribute to overall association test and to the test of maternal transmissions, but not to that of paternal transmissions) and (4) confounding effects TDT-based method to study POE is sensitive to a possible confounding impact of maternal effects on non-equivalence of parental transmissions [, ], as the two cannot be clearly distinguished with this test. Such methodology together with insufficient correction for multiple testing and rather modest sample sizes makes it difficult to draw firm conclusions. In fact, the outcomes of some of these studies on ADHD have either been subject of much controversy [, ] or have failed to replicate [–].One method to account for both maternal effects and non-equivalence of parental alleles is the application of a log-linear modeling in trio data, that can be implemented in a number of softwares (e.g. EMIM/PREMIM, HAPLIN) [, , –]. A likelihood-based approach provides a valid substitute to TDT-based method by performing stratification on both the parental mating type and the inherited number of copies of the allele under study [], allowing for reliable testing of both imprinting and maternal effects. In addition, the likelihood ratio test (LRT) of the log-linear modeling performs better than TDT under dominant or recessive models [, ]. LRT also allows for possible prenatal maternal genetic effects [, ] and incomplete triads []. It is not only robust in detection of POE, but is also able to provide insight into distinguishable imprinting and maternal effects. Furthermore, incorporation of parental origin into association studies may lead to detection of overall genetic effect of an allele (POE and association), possibly incrementing the power of a GWA study and, thus, reducing the sample size needed to observe an association signal, although the required sample size, most likely, still needs to be larger than the ones examined so far.To date, log-linear modeling has not been applied to analyze POE in ADHD. […]

Pipeline specifications

Software tools PLINK, Haplin
Application GWAS
Diseases Genetic Diseases, Inborn