Skip to content

Latest commit

 

History

History
91 lines (57 loc) · 3.79 KB

chap4-data-visualization.md

File metadata and controls

91 lines (57 loc) · 3.79 KB
Raw

Integrative data visualization

  1. Motivation for Vis
  2. What kind of data
  3. Desired feaures of Hi-C Vis tools
  4. Current plights
  5. Overview of tools

Here we'll mainly focused on Hi-C data visulazation, while most of the tools designed for Hi-C data are also able to integragate other 3C-based data to improve the interpretability.

We'll first walk through the motivation for visualizing complex dataset and then discuss the object, demand, current limitations in Hi-C data visualization. Then we'll provide a roadmap with some comparison of some of the most popular softwares available, based on some criteria that people may refer to. Finally we'll choose GIVE and HIGlass as showcases to prepare folk for practical analysis.

1.Motivation for Vis:

  • Explore data feature, find patterns
  • Interpretate the data
  • Verify or raise hypothesis

2.What kind of data:

i) Hi-C data:

  • What is hi-c data, brief intro. may refer to other chapter

###ii) Other useful types of data:

  • genome, epigenome
  • SNP
  • CTCF, ChiP-seq
  • ChIA-PET
  • ...

3.Desired feaures of Hi-C Vis tools:

*** To take a cruise on a Hi-C map is just like navigate on a google map: we want to zoom in & out anytime we want, we want to focus on a special location, get relative information about that location... In a word, we're immersed in the ocean of data but we still want to be as free as a fish.***

  • Multi-omics data, multi-techs data. Optional for users
  • Vivid presentation: Circos, Heatmaps, Birdview.
  • Interactive presentation: mouse zoom-in & out, panning, sliding.
  • Easy to share the data:
    • Cloud based, docker based, block chain(for server), URL share, QR code.
    • Records for eaxct parameters that enable collaborators and public to repeat the trials.
  • Easy for comparison:
  • Compare own data with published papaer —— preloaded dataset.
  • Compare data from different analysis pipelines.
  • Compare 2 datasets in one screen, with exact same scalability and same region, which needs fast and intelligent localization.
  • Desktop applicaton, off-line usage.
  • 3D ,even 4D vis.
  • One stop service, the only desidered input is the raw reads data :D

4.Current plights

  • Size: trillions of pixels of one high resolution heatmap.
  • Resolution at different scale:
    • Compartments:1-10Mb 100-500 Kb bins
    • TADs: 0.1-1Mb 40-100Kb bins
    • Loops: 0.01-0.1Mb 10-40Kb bins
  • Storage of the data, need a standard format ".hic"
  • Smart computation with the need of in time zoom-in & out
  • Input format not uniform. A contact matrix, compressed TAD data or others.
  • Normalization methods for comparing different datasets.
  • ...

5.Overview of tools

Referrence

[1] R. Calandrelli, Q. Wu, J. Guan, and S. Zhong, “GITAR : An open source tool for analysis and visualization of Hi-C data,” 2018.

[2] B. Informatics, “HiGlass : Web-based Visual Exploration and Analysis of Genome Interaction Maps,” pp. 1–31, 2017.

[3] D. Yang et al., “3DIV: A 3D-genome Interaction Viewer and database,” Nucleic Acids Res., vol. 46, no. D1, pp. D52–D57, 2018.

[4] J. T. Robinson, D. Turner, N. C. Durand, H. Thorvaldsdóttir, J. P. Mesirov, and E. L. Aiden, “Juicebox.js Provides a Cloud-Based Visualization System for Hi-C Data,” Cell Syst., vol. 6, no. 2, p. 256–258.e1, 2018.

[5] M. N. Djekidel, M. Wang, M. Q. Zhang, and J. Gao, “HiC-3DViewer: a new tool to visualize Hi-C data in 3D space,” Quant. Biol., vol. 5, no. 2, pp. 183–190, 2017.

[6] G. G. Yardimci and W. S. Noble, “Software tools for visualizing Hi-C data,” Genome Biol., vol. 18, no. 1, 2017.