Extended Data Fig. 1: Correlations between BitAge and StochasticClock.
a) Distribution of transcriptomic age predictions. The 67 neurons of the Calico day 1 dataset were predicted with BitAge and sorted by their predicted age. The x-axis shows the rank of the prediction is ascending order, the y-axis the predicted age. b) BitAge predictions on the CeNGEN and the Calico day 1 dataset are highly correlated (Pearson correlation 0.66, p-value 1.2e-09). The x-axis shows the BitAge predictions of the CeNGEN dataset, the y-axis BitAge predictions of the Calico day 1 dataset. 67 neurons are plotted. The regression model fit with a 95% confidence interval (shadowed area) is shown. c) Stochastic Clock predictions on the CeNGEN and the Calico day 1 dataset are highly correlated (Pearson correlation 0.82, p-value 3.2e-17). The x-axis shows the Stochastic Clock predictions of the CeNGEN dataset, the y-axis BitAge predictions of the Calico day 1 dataset. 67 neurons are plotted. The regression model fit with a 95% confidence interval (shadowed area) is shown. d) Age prediction of BitAge and the Stochastic Clock on Calico Day 1 data. The x-axis shows the predictions by BitAge, the y-axis the prediction by the stochastic clock. There is a strong correlation (Pearson correlation 0.74, p-value 7.8e-13). e) Age prediction of BitAge on CeNGEN bulk data vs CeNGEN pseudobulk. There is a strong correlation (Pearson correlation 0.65, p-value 2.2e-5). f) Age prediction of the Stochastic Clock on CeNGEN bulk data vs CeNGEN pseudobulk. There is a strong correlation (Pearson correlation 0.7, p-value 1.7e-6). g) Predictions with BitAge and the Stochastic Clock on the CeNGEN dataset are highly correlated. The x-axis shows the BitAge predictions of the CeNGEN dataset, the y-axis stochastic data-based clock predictions of the CeNGEN dataset. All 128 neurons are plotted, the 20% youngest and oldest according to BitAge are coloured in blue, the 20% youngest and oldest according to the StochasticClock are coloured in red, overlap between both violet. The remaining neurons are indicated in grey. The regression model fit with a 95% confidence interval (dashed, black lines) is shown. The five neurons the farthest away from the regression line (both, above and below) are indicated. h) BitAge (blue) and StochasticClock (red) predictions on an independent dataset of neuronal RNA from adult day 1 and adult day 8 old nematodes. i) Correlation of median gene length and predicted age of individual neuron types. BitAge clock used on the CeNGEN dataset. Linear fit (orange line) with 95% confidence interval (dashed, orange lines) is shown. Exemplary, young predicted neurons I2, OLL, and PHC (blue); and old predicted neurons ASI, ASJ, and ASK (orange) are indicated. j) Correlation of median gene length and predicted age of individual neuron types. Stochastic clock used on the CeNGEN dataset. Linear fit (orange line) with 95% confidence interval (dashed, orange lines) is shown. k) Correlation of median gene length and predicted age of individual neuron types. BitAge clock used on the day 1 Calico dataset. Linear fit (orange line) with 95% confidence interval (dashed, orange lines) is shown. l) Correlation of median gene length and predicted age of individual neuron types. Stochastic clock used on the day 1 Calico dataset. Linear fit (orange line) with 95% confidence interval (dashed, orange lines) is shown.
