https://github.com/kundajelab/fastISM/raw/docs/images/logo.jpeg

fastISM DocumentationΒΆ

A Keras implementation for fast in-silico saturated mutagenesis (ISM) for convolution-based architectures. It speeds up ISM by 10x or more by restricting computation to those regions of each layer that are affected by a mutation in the input.

Contents:

  • Quick Start
    • Installation
    • Usage
    • Benchmark
    • Getting Help
    • Citation
  • Examples
    • Alternate Mutations
    • Alternate Ranges
    • Multi-input Models
    • Recursively Defined Models
  • Tutorial
    • Installations and Data
    • Init
    • Load Model
    • Benchmark
    • Run on Custom Sequences
    • Sequence Features for a Single Output
  • How it Works
  • Supported Layers
    • Local Layers
    • See Through Layers
    • Aggregation Layers
    • Stop Layers
    • Pooling Layers
  • API
    • ism_base module
    • fast_ism module
    • fast_ism_utils module
    • change_range module
    • flatten_model module
  • Change Log
    • Unreleased
    • 0.5.0 - 2022-02-08
    • 0.4.0 - 2020-09-16
    • 0.3.0 - 2020-08-24
    • 0.2.0 - 2020-08-22
    • 0.1.3 - 2020-08-21
  • Citation

fastISM

Navigation

Contents:

  • Quick Start
  • Examples
  • Tutorial
  • How it Works
  • Supported Layers
  • API
  • Change Log
  • Citation

Related Topics

  • Documentation overview
    • Next: Quickstart

Quick search

©2022, Kundaje Lab. | Powered by Sphinx 1.8.6 & Alabaster 0.7.12 | Page source