Skip to content

README.md

Latest commit

 

History

History
111 lines (81 loc) · 4.57 KB

README.md

File metadata and controls

111 lines (81 loc) · 4.57 KB
Raw

Information Module

Information-theoretic analysis for biological sequences and multi-omic data, from Shannon entropy to Fisher-Rao geometry.

Overview

Information-theoretic analysis for biological sequences and multi-omic data, from Shannon entropy to Fisher-Rao geometry.

Table of Contents

Architecture

graph TD
    subgraph "Information Module"
        MC[metrics/core/] --> |syntactic.py| SH[Shannon Entropy, MI, KL]
        MC --> |continuous.py| DE[Differential Entropy]
        MC --> |estimation.py| EST[Estimators]

        MA[metrics/advanced/] --> |geometry.py| FG[Fisher-Rao, Hellinger]
        MA --> |channel.py| CH[Channel Capacity]
        MA --> |decomposition.py| DC[PID, Synergy]
        MA --> |semantic.py| SEM[Semantic Similarity]

        MN[metrics/analysis/] --> |analysis.py| PRO[Information Profiles]

        NI[network_info/] --> |information_flow.py| TF[Transfer Entropy, Granger]

        INT[integration/] --> |integration.py| BI[Cross-Omic Integration]

        WF[workflow/] --> |workflows.py| BAT[Batch Pipelines]
    end
Loading

Key Capabilities

Core Measures

Function Module Description
shannon_entropy metrics.core.syntactic Shannon entropy from probability distributions
mutual_information metrics.core.syntactic Mutual information between discrete variables
kl_divergence metrics.core.syntactic Kullback-Leibler divergence
conditional_entropy metrics.core.syntactic H(X|Y) conditional entropy
transfer_entropy metrics.core.syntactic Directed information transfer with lag
differential_entropy metrics.core.continuous Entropy of continuous distributions

Advanced Measures

Function Module Description
fisher_rao_distance metrics.advanced.geometry Geodesic distance on statistical manifolds
channel_capacity metrics.advanced.geometry Blahut-Arimoto channel capacity
information_bottleneck metrics.advanced.geometry Information bottleneck compression
hellinger_distance metrics.advanced.geometry Hellinger distance between distributions
jensen_shannon_divergence metrics.core.syntactic Symmetric divergence measure

Submodules

Module Purpose
metrics/ Information-theoretic measures organized into core/ (syntactic, continuous, estimation), advanced/ (geometry, channel, decomposition, semantic), and analysis/ (information profiles)
network_info/ Transfer entropy, Granger causality, information flow networks
integration/ Cross-omic information integration (DNA, RNA, single-cell)
workflow/ Batch entropy analysis and end-to-end pipelines

Quick Start

from metainformant.information.metrics.core.syntactic import shannon_entropy, mutual_information, kl_divergence
from metainformant.information.metrics.core.continuous import differential_entropy
from metainformant.information.metrics.advanced.geometry import fisher_rao_distance
from metainformant.information.workflow.workflows import batch_entropy_analysis

# Shannon entropy of a nucleotide distribution
H = shannon_entropy([0.3, 0.2, 0.25, 0.25])

# Mutual information between two genomic features
mi = mutual_information(feature_x, feature_y)

# Fisher-Rao geodesic distance between distributions
dist = fisher_rao_distance([0.5, 0.3, 0.2], [0.4, 0.4, 0.2])

# Batch entropy analysis across sequences
results = batch_entropy_analysis(sequences=["ATCGATCG", "GCTAGCTA"], k=2)

Integration

  • DNA -- Sequence entropy profiles via batch_entropy_analysis
  • RNA -- Expression variability measured with differential_entropy
  • Networks -- Information flow with transfer_entropy and Granger causality
  • ML -- Feature selection using mutual_information and information_coefficient

Related