Updated documentation to account for new default model.

README.md

@@ -44,3 +44,18 @@ If you use our code in your research, please cite the following paper:
 L. Theis, P. Berens, E. Froudarakis, J. Reimer, M. Roman-Roson, T. Baden, T. Euler, A. S. Tolias, et al.  
 [Supervised learning sets benchmark for robust spike detection from calcium imaging signals](http://bethgelab.org/publications/127/)  
 bioRxiv, 2014
+
+The default model was trained on many datasets (together containing roughly 110,000 spikes) from
+different labs. Therefore, if you use the default model for prediction, please also cite:
+
+J. R. Cotton, E. Froudarakis, P. Storer, P. Saggau, and A. S. Tolias  
+Three-dimensional mapping of microcircuit correlation structure  
+Frontiers in Neural Circuits, 2013
+
+J. Akerboom et al.  
+Optimization of a GCaMP calcium indicator for neural activity imaging  
+Journal of Neuroscience, 2012
+
+T. W. Chen et al.  
+Ultrasensitive fluorescent proteins for imaging neuronal activity  
+Nature, 2013

doc/scripts/tutorial.rst

@@ -141,9 +141,9 @@ preprocessed yet, use
 
 The predictions are saved in the same format as the data files, except that the entries
 ``spikes``, ``spike_times`` and ``calcium`` are removed to save space. By default, the prediction
-uses a model which has been trained on data recorded from V1 of mice using OGB1 as indicator. But
-it is possible to train a model which is better adapted to our data. Once trained, the model can be
-used for prediction as follows:
+uses a model which has been trained on several datasets recorded by different labs under different
+conditions. These datasets combined contained roughly 110,000 spikes. But it is possible to train
+a model specifically for our data. Once trained, the model can be used for prediction as follows:
 
 .. code-block:: bash
 
@@ -171,6 +171,7 @@ To print a list of available parameters to influence the training, please see:
 
     $ c2s train -h
 
+
 Evaluation
 ----------