Genome-wide association study on electrophysiological correlates of human performance monitoring
Final Report Abstract
The goal of our project was to search for associations between genetic polymorphisms and electrophysiological endo-/phenotypes of human action monitoring in a genome-wide study. We therefore recorded EEG from a large sample of healthy participants while they were working on different classical performance monitoring paradigms (e.g., error- or feedback-processing) to derive phenotypes for further genetic association analyses. While preparing the EEG phenotypes for further genetic analysis, we were able to derive a number of interesting conclusions from our sample. We were able to show that male subjects showed a higher ERN amplitude but that the dynamics of coupling between ERN and post-error slowing between men and women was comparable. Errors in human performance might trigger a number of different behavioral adaptation processes meant to lower the probability of future errors. Fitted model parameters and their independently measured neuronal proxies in beta power convergently showed a complex interplay of multiple mechanisms initiated after mistakes. Suppression of distracting evidence, response threshold increase, and reduction of evidence accumulation caused slow and accurate post-error responses. This provides evidence for both adaptive control and maladaptive orienting after errors yielding an adaptive net effect. Often the question remains open on how to best quantify these adaptive behavioral effects after an error. To obtain unbiased PES scores for interference tasks, we proposed to compute unweighted individual-participant means by initially calculating mean RTs for congruent and incongruent trials separately, before averaging congruent and incongruent mean RTs to calculate means for post-correct, pre-error and post-error trials. These considerations related to the question of the overall robustness of the ERN given different numbers of error trials between subjects. We were able to show that across participants, the number of errors correlated with the amplitude of the ERN independently of the number of errors included in ERN quantification per participant, constituting a possible confound when such variance is unaccounted for. Additionally, we found that ERN amplitudes reach high consistency within participants at lower trial numbers, yet when comparisons between groups of participants are desired, increasing error-trial numbers lead to higher statistical power. With respect to feedback processing first results supported the theory of the FRN as a representation of a signed RPE. Additionally, our data indicated that surprising positive feedback enhanced the EEG response in the time window of the P3. These results corroborated previous findings linking the P3 to the evaluation of PEs in decision making and learning tasks. In this project two novel analysis methods for the given genetic data situation have been presented: a univariate combined approach trying to bridge the gap between nearly disjunct SNP datasets, and the SIR method as a more intuitive alternative to pICA. . Independent of method usage, it was shown clearly that disjoint datasets lead to weaker result and should be avoided in study planning. Using different methods multiple genes were identified as candidates with potential influence on FRN amplitude. TMEM132C, RBFOX1, STIM1, SNRPB, and CSMD1 as well as a potential association of nongene specific SNPs are implicated as possibly relevant. Gene expression of all those genes is especially high in brain tissue, which would be in favor of the importance of these genes in action monitoring. Further genetic analyses are currently running using alternative phenotypes and the power of the whole sample. In the second project phase we were largely impacted by the corona pandemic which resulted in a substantial delay in data acquisition. Furthermore, the necessary merging of genetic data based on different analysis systems proved to be time-consuming and suboptimal for various reasons.
Publications
-
Gender Influences on Brain Responses to Errors and Post-Error Adjustments. Scientific Reports, 6(1).
Fischer, Adrian G.; Danielmeier, Claudia; Villringer, Arno; Klein, Tilmann A. & Ullsperger, Markus
-
Comparing the error‐related negativity across groups: The impact of error‐ and trial‐number differences. Psychophysiology, 54(7), 998-1009.
Fischer, Adrian G.; Klein, Tilmann A. & Ullsperger, Markus
-
Cortical beta power reflects decision dynamics and uncovers multiple facets of post-error adaptation. Nature Communications, 9(1).
Fischer, Adrian G.; Nigbur, Roland; Klein, Tilmann A.; Danielmeier, Claudia & Ullsperger, Markus
-
Unbiased post-error slowing in interference tasks: A confound and a simple solution. Behavior Research Methods, 54(3), 1416-1427.
Derrfuss, Jan; Danielmeier, Claudia; Klein, Tilmann A.; Fischer, Adrian G. & Ullsperger, Markus
-
Disentangling performance-monitoring signals encoded in feedback-related EEG dynamics. NeuroImage, 257(2022, 8), 119322.
Kirsch, Franziska; Kirschner, Hans; Fischer, Adrian G.; Klein, Tilmann A. & Ullsperger, Markus
