blueberry¶
blueberry implements efficient data structures for the storage, analysis, and exploration of Hi-C datasets. Currently it supports data types for the storage of raw Hi-C datasets, performing KR matrix-balancing, observed/expected normalization, calculation of correlation matrices, and calculation of the first eigenvector, as well as datatypes for Fit-Hi-C results.
This page currently serves as an API reference for these data structures.
Contents:
Data Structures¶
blueberry currently implements two major data structures, the ContactMap, and the FithicContactMap. The ContactMap is intended for data files which have the format i, j, contactCount. The FithicContactMap is intended for data files which have the format i, j, contactCount, p, q.