First Connectomics Challenge: From Imaging to Connectivity

Abstract

We organized a Challenge to unravel the connectivity of simulated neuronal networks. The provided data was solely based on fluorescence time series of spontaneous activity in a network constituted by 1000 neurons. The task of the participants was to compute the effective connectivity between neurons, with the goal to reconstruct as accurately as possible the ground truth topology of the network. The procured dataset is similar to the one measured in in vivo and in vitro recordings of calcium fluorescence imaging, and therefore the algorithms developed by the participants may largely contribute in the future to unravel major topological features of living neuronal networks from just the analysis of recorded data, and without the need of slow, painstaking experimental connectivity labeling methods. Among 143 entrants, 16 teams participated in the final round of the challenge to compete for prizes. The winners significantly outperformed the baseline method provided by the organizers. To measure influences between neurons the participants used an array of diverse * The two first authors contributed equally.

Invasions of Host-Associated Microbiome Networks

“…Once sufficiently long time-series can be measured, comparisons of cross correlations of the abundances from microbiome communities across body sites could provide signatures of leaders and laggards in the whole body microbiome dynamics. Finally, we mention that network reconstruction competitions based on benchmark datasets can be a fruitful way to attract highperforming algorithms (e.g., Orlandi et al 2014). …”

Inferring Functional Neural Connectivity with Deep Residual Convolutional Networks

“…In parallel, several alternative methods for inferring functional connectivity have been published. One method, using a partial correlation coefficient metric estimated from the inverse covariance matrix (Sutera et al, 2014), set the state-of-the-art benchmark at the inaugural Kaggle ChaLearn Connectomics competition, edging out baseline summary statistics like correlation coefficients and entropy-based causality estimations (Orlandi et al, 2014).…”

“…Synthetic Graph Architecture. The Kaggle ChaLearn Connectomics data are generated from a realistic model of neural dynamics (leaky integrate-and-fire), calcium binding, and fluorescence (Stetter et al, 2012;Orlandi et al, 2014). Graph model parameters were tuned so that cells were spontaneously active, including periods of pan-graph bursting.…”

“…Nevertheless, there have been attempts to extract meaningful connectivities from these recordings. Recently, several methods were published together as part of the Kaggle ChaLearn Connectomics competition (Orlandi et al, 2014). Here, we report an improved deep learning method that outperforms the competition leader and, importantly, provides a viable false discovery rate (FDR) at a true positive rate (TPR) poised to reveal millions of synaptic connections in real brain volumes.…”

Super-selective reconstruction of causal and direct connectivity with application toin-vitroiPSC neuronal networks

“…Only a few techniques can compete in the challenge of inferring the effective connectivity of a network. Partial-correlation [31,47], which takes into account all neurons in the network, showed best performance in detecting direct associations between neurons and filtering out spurious ones [48]. The most significant limitation of this solution is its high computational cost.…”