Gene Expression Analysis

Data Source

The dataset was aggregated from many source files from a publicly available database. https://portal.gdc.cancer.gov/

The project ID used was TCGA-OV, which studies specifically Ovarian Serous Cystadenocarcinoma.

Project Navigation

• csv
  • data.csv | raw complimation of every patient and every data field
  • SurvivalNormal.csv | normalized ovarian cancer patients with just gene fields and survival outcome
  • SurvivalNormalBoth.csv | normalized ovarian and breast cancer patients with gene fields, age, cancer information, and prognosis.
  • ProgressionNormal.csv | normalized ovarian cancer patients with gene fields and progression outcome
  • RecurrenceNormal.csv | normalized ovarian cancer patients with gene fields and recurrence outcome
• R_Analysis
  • analysis.Rmd | R markdown script that conducts analysis of dataset, creating different types of models
  • gen_csv.Rmd | R markdown script that generates SurvivalNormalBoth.csv from data.csv
  • control_analysis.Rmd | R markdown script that conducts same analysis of dataset as analysis.Rmd however it substitutes the actual data for randomly generated data to serve as a control.
  • control_trials.py | Python script to run the R markdown script multiple times and average the results. It's use-case is python control_trials.py 100 where the argument is the number of trials to run.
• Hiearchical_Clustering
  • hierarchical_clustering.py | python script that generates the figures for hierarchical clustering
• NeuralNetwork
  • choosing_genes.py | the python script for selecting the optimal features
  • graphing_feature_selection.py | python script that graphs the results of choosing_genes.py
  • graphing_neural_performance.py | python script that graphs the performance of the neural networks
  • neural_model.py | python script that includes functions for building a model on the dataset. Also includes a general evaluation of the a model
  • SurvivalExperiment.txt | the results of choosing_genes.py optimizing survival predictions
  • ProgressionExperiment.txt | the results of choosing_genes.py optimizing progression predictions
  • RecurrenceExperiment.txt | the results of choosing_genes.py optimizing recurrence predictions
• Learning_Curve
  • graphing_learning_curve.py | graphs the learning curves from output files (hard-coded names)
  • learning_curve.py | generates a learning curve datafile on neural network
  • learning_curve.Rmd | generates a learning curve datafile on logistic regression model
• README.md | a project description
• LICENSE | the license for this project

Dataset Fields

General Info

Patient ID, Cancer Type, Survival Status, Survival in Days (7300 represents continued survival), Progression Status, Recurrence Status, Age, Race

Genes

XRCC6, XRCC5, XRCC7, LIG4, LIG3, LIG1, XRCC4, NHEJ1, XRCC1, DCLRE1C, TP53BP1, BRCA1, BRCA2, EXO1, EXD2, POLM, POLL, POLQ, RAD50, MRE11, NBN, TDP1, RBBP8, CTBP1, APLF, PARP1, PARP3, PNKP, APTX, WRN, PAXX, RIF1, RAD52, RAD51, ATM, ATR, TP53, H2AX, ERCC1, ERCC4, RPA1, MSH2, MSH3, RAD1, MSH6, PMS1, MLH1, MLH3

Tumor Information

ICD10 Code, Figo Stage, Neoplasm Histologic Grade, Tumor Size, Lymphatic Invasion Status

Treatment Information

Total Dose, Number of Treatment Cycles, Treatment Day Range, Therapy Type, Drug Name, Regimen Indication

Null Fields

Fields are null because they are either missing, unavailable, or irrelevant to the original study.

Dataset Composition

• Ovarian Count: 374
• Breast Count: 279
• Ovarian Count Alive: 145 Deceased: 229
• Breast Count Alive: 137 Deceased: 142
• Mean: Ovarian Length Survival (yr): 9.741966285945132
• Mean: Breast Length Survival (yr): 12.21966448282708
• Standard Deviation: Ovarian Length Survival (yr): 8.349660685645034
• Standard Deviation: Breast Length Survival (yr): 8.12829931819985
• Mean: Ovarian Age (yr): 59.58288770053476
• Standard Deviation: Ovarian Age (yr): 11.343556969016111

Ovarian Cancer Aggregated Figo Stage

Type	Total	Percentage
Stage IIIC	271	0.72460
Stage IIC	15	0.04011
Stage IIB	3	0.00802
Stage IV	57	0.15241
Stage IIIA	7	0.01872
UNKNOWN	3	0.00802
Stage IIIB	14	0.03743
Stage IC	1	0.00267
Stage IIA	3	0.00802

Breast Cancer Aggregated Figo Stage

Type	Total	Percentage
UNKNOWN	279	1.00000

Ovarian Cancer Aggregated Race

Type	Total	Percentage
white	324	0.86631
asian	11	0.02941
african american	25	0.06684
UNKNOWN	11	0.02941
american indian or alaska native	2	0.00535
native hawaiian or other pacific islander	1	0.00267

Breast Cancer Aggregated Race

Type	Total	Percentage
UNKNOWN	279	1.00000

Ovarian Cancer Aggregated ICD 10 Code

Type	Total	Percentage
C56.9	369	0.98663
C48.1	4	0.01070
C48.2	1	0.00267

Breast Cancer Aggregated ICD 10 Code

Type	Total	Percentage
UNKNOWN	279	1.00000

Ovarian Cancer Aggregated Neoplasm Histologic Grade

Type	Total	Percentage
G3	320	0.85561
G2	42	0.11230
G1	1	0.00267
GB	2	0.00535
GX	6	0.01604
UNKNOWN	2	0.00535
G4	1	0.00267

Breast Cancer Aggregated Neoplasm Histologic Grade

Type	Total	Percentage
UNKNOWN	279	1.00000

Ovarian Cancer Aggregated Tumor Residual Disease

Type	Total	Percentage
Microscopic	64	0.17112
11-20 mm	26	0.06952
UNKNOWN	43	0.11497
1-10 mm	171	0.45722
>20 mm	70	0.18717

Breast Cancer Aggregated Tumor Residual Disease

Type	Total	Percentage
UNKNOWN	279	1.00000

Ovarian Cancer Aggregated Lymphatic Invasion

Type	Total	Percentage
NO	47	0.12567
UNKNOWN	227	0.60695
YES	100	0.26738

Breast Cancer Aggregated Lymphatic Invasion

Type	Total	Percentage
UNKNOWN	279	1.00000

Progression Aggregated

Type	Total	Percentage
FALSE	222	0.59358
TRUE	100	0.26738
UNKNOWN	52	0.13904

Recurrence Aggregated

Type	Total	Percentage
FALSE	190	0.50802
TRUE	132	0.35294
UNKNOWN	52	0.13904