This repository contains three sets of codes to perform analysis of the hemibrain dataset presented in Tanaka & Clark (2021) bioRxiv as follows:
(1) Identification of putative T5 neurons (identify_T5)
(2) Clustering of lobula columnar inputs into LPLC1 (lobula_clustering)
(3) Identification of downstream pathways of LPLC1 (downstream_analysis)
The scripts here interface with the hemibrain dataset using neuprint-python. To access the dataset, you need to sign up on neuPrint+ (Google account required). Find your Auth Token from the Account menu on neuPrint+, and save it under this directory as a plane text file named authtoken (without extension). The scripts read this authtoken file to connect to the hemibrain database. Other than neuprint-python, the scripts require common packages like numpy, scipy, pandas, matplotlib, sklearn etc.