1 of 13

BigPapi

Making renal papillary cancer research more accessible and reproducible through Google BigQuery.

StudenTec team

2 of 13

BigQuery and Jupyter

Used Google’s Genomics API to load genomic data files into BigQuery

Created reproducible SQL queries against TCGA data to identify any patterns.

R Notebook to serve as starting point

3 of 13

DATA EXPLORATION

https://github.com/SVAI/studentec/blob/master/experiments.ipynb

4 of 13

Using Patient (Bill’s data) via BigQuery (4 hits)

5 of 13

Non Synonymous SNP HITS on KIRP data:

ST6GALNAC5 - Domain

FLG2

AMPD2 - Domain

DPP6

6 of 13

ST6GALNAC5

7 of 13

8 of 13

9 of 13

10 of 13

AMPD2

11 of 13

AMPD2 - Adenosine Monophosphate Deaminase 2

The protein encoded by this gene is important in purine metabolism by converting AMP to IMP. The encoded protein, which acts as a homotetramer, is one of three AMP deaminases found in mammals.

RNA Seq data from TCGA

12 of 13

Non synonymous hit on Chr 5: FTMT

Query on all SNP deletions - frameshift on FTMT which is involved in hipoxia regulation.

Ferritin is an iron storage protein. Iron is involved in Wnt signalling. Wnt can do things like make hair grow back.

https://www.nature.com/articles/1210711 - A role for iron in Wnt signalling

13 of 13

Conclusions & Learning Experience

EVERYONE SHOULD BE USING BIGQUERY FOR GENOMIC DATA ANALYSIS

SCIENCE NEED MORE HACKATHONS