1 of 12

a data repository and analysis tool for quantifying metabolism in vivo

Michael Neinast

Lance Parsons

Fan Kang

Robert Leach

John Matese

designed for infusions

2 of 12

Tracebase is designed for steady-state infusion experiments

designed for infusions

3 of 12

Upload data from any infusion!

From one infusion

From many

Compare anything

4 of 12

Catheterized mice infused with isotope-labeled nutrient

Mass-spectrometry measures compounds and their isotope labeling

.raw

.mzXML

Raw data files

Researchers extract the abundance and labeling of compounds

.xlsx

experiment information

results

Experiment Workflow

5 of 12

TraceBase optimizes data for any analysis.

Consistency

  • resolves naming discrepancies
    • compounds, tissues, genotypes…
  • uniform calculations
  • “tidy” format: 1 row = 1 observation

Ready for variety of other tools

  • minimal code-to-answer
  • data browsing in ShinyApp

PeakData.tsv

PeakGroups.tsv

Fcirc.tsv

6 of 12

What can you do with TraceBase?

7 of 12

Browse experiments and results.

8 of 12

Download -> Compare Data across datasets

~30 minutes in R

9 of 12

Uniform organization -> compare anything

Mike MacArthur

PeakData.tsv

PeakGroups.tsv

Fcirc.tsv

10 of 12

Use PeakData to go deeper

Ratio of M+3 : M+2 labeling in malate from glucose infusions

Use everything to model fluxes

Daniel Weilandt

11 of 12

Mine raw data to find see more compounds.

Xi Xing

.mzXML

download well-organized mzXML

mine data for additional compounds which the original researcher had ignored.

12 of 12

How could TraceBase work outside of our lab?

Rabinowitz Lab

Jang?

Hui?

TeSlaa?

Reynolds?

Bartman?

Public

Private instances of TraceBase hosted by each lab

Public instance for published data only

submitted by anyone

hosted by LSI at Princeton

Arany?

Deberardinis?

Rathmell?