hms-dbmi / OncoThreads

OncoThreads longitudinal cancer genomics visualization project.
http://oncothreads.gehlenborglab.org
MIT License
10 stars 1 forks source link

Walk-through tutorial #252

Closed wangqianwen0418 closed 3 years ago

wangqianwen0418 commented 3 years ago
wangqianwen0418 commented 3 years ago

Tutorial Steps

Overview

  1. tabbed panel image
    
    To start with, select different views to analyze the clinical sequences from different aspects.

Block View groups patients at each tiempoint based on their values of one selected feature.

State Transition provides a more advanced analysis and enables state identification using timepoint features.

Timeline View shows the individual clinical sequence of each patient.


State Transition Tab
1. state identification
<image src="https://user-images.githubusercontent.com/19774198/102569929-563c1080-40b4-11eb-947b-710bfd3e1402.png" width="500"/>

Step 1: State Identification In the Scatter Plot, each point indicates the feature values of one patient at one timepoint. Different color indicates different states. You can draw a lasso to modify the identified states or directly change the number of states in the top left input box.

2. transition overview
<image src="https://user-images.githubusercontent.com/19774198/102570114-bc289800-40b4-11eb-8b8d-4a53c9c7b42f.png" width="500"/>

Step 2: analyze the state transition among all patients. The y-axis presents the timeline and the colored rectangle indicates patients of the same state. You can group patients based on their state transitions by changing the number in the top left input box.


3. frequent patterns

<image src="https://user-images.githubusercontent.com/19774198/102570305-10337c80-40b5-11eb-9238-51ec00aa3160.png" width="400" />

Step 2: analyze the state transition among all patients. This table summarizes the frequent state transition patterns. You can sort the rows or search frequent patterns by clicking the icons in the table header.


4. detail view
<image src="https://user-images.githubusercontent.com/19774198/102570421-4d980a00-40b5-11eb-86be-e64f5037c08b.png"  width="500" />

Step 3: Detailed Analysis You can select interested patient groups and observe the state transition details.


5. dropdown menu and feature manager to add more time-point features/events

![image](https://user-images.githubusercontent.com/19774198/102570499-77513100-40b5-11eb-867d-bd23439bcb09.png)

Add more features through the drop down menu and the Feature Manager.

(ง •_•)ง Having Fun with your exploration!

wangqianwen0418 commented 3 years ago

Meeting feedback