3a5ranvhbbo.html

<div></div><ul><li>The combination of CsChrimson and GCaMP does not strictly insure that the responses observed are direct. We did choose to use low stimulation intensities to limit indirect effects (see figure SXX). The sparseness of the connectivity matrix obtained and control experiments both suggest that this is a reasonable assumption. Small, unreliable responses were interpreted to be the result of indirect effects (see next pargraph).&nbsp;</li><li>The sensitivity of the method is bounded by the stimulation protocol and the sensitivity of GCaMP6-m. This means that an absence of response cannot be formally be interpreted as an absence of connection. Given the range of responses observed, the most difficult pairs to interpret are the ones generating small or unreliable responses upon stimulation.</li><li>The full-field stimulation protocol used here can potentially lead to network artifacts. This is another reason for us to have used low intensity stimulations. Stimulating the full presynaptic population could lead to the recruitment of large inhibitory feedbacks, or other network modules. When we suspected this could be a possibility, we tried to control for it using pharmacology, and never saw evidence of "hidden responses"(<b>see Discussion</b>).</li><li>In the ex vivo preparation used, the network state is not controlled. It is possible that some connections are functionally conditional on the overall state of the network (<b>ref Allan/Reiser</b>). The fact that we observe large baseline fluctuations however would tend to indicate that a variety of network states were reached during the experiments.&nbsp;</li></ul>