FAQ about Beat synchronization GWAS study

This page is an FAQ about Niarchou, Gustavson et al., 2022, Nature Human Behaviour. Genome-wide association study of musical beat synchronization demonstrates high polygenicity.

Access the full paper at https://doi.org/10.1038/s41562-022-01359-x

Q: What are the study’s main findings?

In this study, we performed a genome-wide association study (GWAS) of beat synchronization in over 600,000 individuals. We discovered that the genetic architecture of this trait is highly polygenic, meaning that it is influenced by many genes in the human genome. We identified 69 separate locations on the genome in which different genetic alleles in the population account for some of the variability in how accurately people synchronize to a musical beat (according to their self-reported beat synchronization skill, which we also validated; see question below). Genes associated with beat synchronization are more likely than chance to be genes involved in central nervous system function, including genes expressed in brain tissue and genes involved in early brain development. We also discovered that beat synchronization shares some of its genetic architecture with other traits, including several that are involved in biological rhythms (walking, breathing, and circadian chronotype).


Q: Is beat synchronization the same as rhythm ability?

A: Beat synchronization, or the degree to which an individual can synchronize their movements accurately in time with a musical beat, is one particular human rhythm trait. Music cognition scientists also study other rhythm-related traits (for example, the accuracy with which an individual can tell if two musical rhythmic sequences are the same or different, or the degree to which an individual can keep a steady beat without a reference metronome). In the present study, we assessed beat synchronization (in the genetic study sample) by asking people if they can clap in time with a musical beat. To be sure that this self-report question is an accurate way to assess people’s beat synchronization, we asked another group of people to tap in time to the beat of musical excerpts, and to perform a rhythm perception task. The results of these “phenotype validation tasks” showed that how people respond to the self-report question is correlated with how accurately they tap to the beat of music (and also to how well they can perceive differences in musical rhythms), thus suggesting that the self-report question is a good proxy for actually measuring people’s beat synchronization and related rhythm abilities.


Q: Is rhythm in your genes?

A: In short, yes, rhythm (beat synchronization) is genetically influenced! The longer, more nuanced answer, though, is that within the population, we were able to see how genetic variants that differ from person to person account for some of the overall variability in beat synchronization. Certain genetic variants were more prevalent in the study group that self-identified as having higher beat synchronization ability versus the group with difficulties with beat synchronization. However, this is far from genetic determinism! The biology of rhythm and of musicality in general is very complex, and humans interact with their environment in complex ways that are not always discoverable with the type of study design that we used. We discuss this idea below in more detail (see the question on our study’s limitations and what does the study not say).


Q: Can we predict someone’s rhythm ability based on their genes?

A: We cannot make definitive predictions at the individual level. One of the methods we used, called polygenic scores, computes the sum of genetic effects associated with beat synchronization in each individual (using weights derived from the primary GWAS dataset), but the resulting polygenic scores are only a rough guess: they can tell us only what an individual’s likelihood of specific levels of beat synchronization would be in relation to the population-based model, but they do not correspond directly to an exact match with the person’s beat synchronization accuracy. A better way to do that would be to test their beat synchronization directly, with a rhythm test! However, we are still excited about the current results because when we pool together data from many individuals, we are able to use these models to start to get a foothold on the underlying biology and to explain a small amount of the variation in the phenotype. As we point out in the Discussion, results from this type of study should be used fairly and ethically, and never for harm (i.e. we should not decide which children should get music training based their genetics; children should have access to music education because it is an important part of our culture and well-being in society).


Q. The paper talks about the enrichment of certain biological functions; what does it mean for the genetic architecture of beat synchronization to be enriched for a biological function? 

A: Here we are talking about a greater than chance likelihood that the genes involved in our phenotype have a particular biological function. When we say that that the genetic architecture of beat synchronization is enriched for genes expressed in brain tissue, or for genes involved in neurodevelopment or synaptic transmission we mean that that the list of genes significantly linked to beat synchronization in the initial GWAS analyses include many genes expressed in brain tissue, and so on. 


Q. What are our study’s limitations and what does our study NOT say?

This is the first very large-scale GWAS of a musicality trait, and as such there were several initial limitations. First, the phenotype in the GWAS is quite simple (“yes” versus “no” responses to the question Can you clap in time with a musical beat?) and future GWAS of rhythm could make use of more detailed rhythm questions, or directly measuring participants’ beat synchronization task performance. Also, the current study was conducted in individuals of European ancestry (this is the genetics term for white people), and while we believe the main results would be broadly similar in other groups, we definitely need to do studies in groups from other genetic ancestries to fully understand the genetic architecture of the trait.

Importantly, the current study only shows that we’ve been able to use genetics to explain a portion of the variability in beat synchronization skills (again, at the level of pooled data in a large study sample). Where we talk about “heritability” we are referring to the amount of phenotypic variance explained by genetic variation. This does not mean that rhythm is only “genetic” versus only “environmental,” or that rhythm is genetic in certain people but not others. We already know from the twin-based literature that rhythm skills, as well as other musicality traits - including even how much music training people pursue -  are in part genetically influenced and in part environmentally influenced. While colloquially we often have intuitions and biases about whether a person’s music accomplishment are caused by their genetics (talent “running in the family”) or by the amount and quality of training (that “10,000 hours” idea), these are in fact biases and assumptions; scientifically we really can’t say for sure how and why an individual reaches (or does not reach) a certain level of musicality. So it’s not “either-or” but “both-and” genes and environment, and the incredibly complex biological interrelationships that occur during human development of musicality will take many, many more years of work to unravel!


Q: Is perfect pitch related to rhythm? 

A: Perfect pitch, known as “absolute pitch” to scientists studying the phenomenon, is the (rare) ability to identify and name musical pitches without use of a reference pitch for comparison. Studies in twins and families (for example, Barharloo et al., 2001) have shown that absolute pitch ability is somewhat heritable (meaning that genes partially influence the trait). We did not examine absolute pitch in our study, but there is not an equivalent in the rhythm domain: people with high musical talent overall may excel at both pitch and rhythm, but we don’t know to what extent those overlaps are genetically driven. Moreover, many individuals with impairments in processing/detecting changes in musical pitch (i.e. congenital amusia) do not demonstrate rhythm impairments (Peretz & Vuvan, 2017). Also, most individuals with high music aptitude when measured with melodic tasks do not have absolute pitch (it is a relatively rare ability). However, we do know that the rhythmic structure of music (including the feeling of a “pulse” or beat) helps orient our attention to specific important moments in musical time, and thus beat can change the way that our brains process melodies. More research is needed to understand the complex genetic and phenotypic relationships between rhythm abilities (including beat synchronization) and pitch-related abilities.


Q: What are the implications of the study? 

A: We believe that this study marks an important step forward for the emerging field of musicality genetics, as we demonstrated that rhythm can be assessed reliably with self-report measures deployed at very large scale in a population cohort, and that the resulting GWAS of beat synchronization shows an expected pattern of polygenicity. Moreover, we found that the genetic architecture of beat synchronization is enriched for genes expressed in brain tissues, and excitingly, in motor and timing areas of the brain such as the basal ganglia and cerebellum that are known from neuroimaging studies to be active while participants perform beat synchronization and beat perception tasks. Also, the findings of enrichment in genes expressed in fetal brain development suggest that the brain might begin to wire itself for sensitivity to musical beat structure very early in neurodevelopment. Considered together, these results have implications for connecting the genetic architecture of beat synchronization to its neural architecture. Finally, we found interesting genetic correlations between beat synchronization and a constellation of interrelated traits: walking pace, musculoskeletal strength, breathing function, and processing speed; the shared genetic architecture has implications for physical and cognitive function during aging. Thus, with this first GWAS of beat synchronization we have paved the way for future basic and clinical-translational work on various aspects of the genetics of rhythm.