Capabilities and Technology

lab panoramic

Since 2008, the Metabolomics Platform has performed over 1000 projects, collaborated with hundreds of research groups, and collected >700,000 sample acquisitions of cells, tissues, plasma, other biofluids and environmental samplings. In doing so, we have developed infrastructure and approaches to perform a variety of targeted and nontargeted metabolomics analyses using mass spectrometry-based approaches.

If you would like to work with us, please consult us prior to designing your experiment. We are happy to collaborate if you have an exciting project and we have available resources. ​​​​

Capabilities

We are equipped to deliver high-quality data for a broad spectrum of applications. Let us know if your requested experiment is not listed below. We employ a team of professional staff, scientists and software engineers who are eager to push the boundaries of metabolomics.

  • Discovery (Nontargeted) - Full spectrum acquisition on HRAM Orbitrap instruments to report both knowns and unknowns. For our specific LCMS methods and our processing workflow, see our Technology section below.
  • Hypothesis Driven (Targeted) - For analytes of interest, we perform targeted extraction using custom software (manuscript in prep) to ensure their presence and accuracy in the repoted feature table. For increased sensitivity, we employ Triple-Quadrupole MS/MS instruments. See our Technology section for more information.
  • Large-Scale Human Population Studies (>2,000 participants) - We have developed infrastructure for sample handling, acquisition, feature extraction, dataset stitching and drift correction for large population studies. Tahir et. al, Nature Communications 2022, Mayers et. al, Nature Medicine, 2015
  • Exposomics and Xenometabolomics - We are developing in-house custom software to identify and extract low abundance/low occurence analytes, and preserve features as batches are merged. We will continue to develop this in collaboration with the Dietary Biomarkers Development Consortium  Chakraborty et. al, Current Developments in Nutrition, 2025
  • Compound Identification and Structural Elucidation - Through informations, comparison to standards, MS/MS, purification, and organic chemistry, we can identify and confirm novel metabolites.
  • Cell-Based Phenotyping - Li et. al, Nature Medicine, 2019
  • Stable Isotope Labeling - Lee et. al, Nature, 2023
  • Pharmacometabolomics and Mechanistic Studies
  • High Throughput Screening (Coming soon)

Each of our capabilities leverages the infrastructure we've built in the time since we were established in 2008.  This includes considerations for:

  • Experimental Design: Study design, sample requirements, handling, and timing to ensure high-quality, interpretable results
  • Sample Processing and Handling: Homogenization, aliquoting, and batching
  • Quality Control Measures: Use of internal standards, pooled-reference technical replicates (PREFs) and tube blanks
  • Acquisition Strategy: Queue design including randomization and QC sample placement
  • Data Processing: Feature extraction, annotation via authenticated MS1/MS2 standards, drift correction, and batch integration

For more on our scientific work, please see this overview of our research and our publication list.

Technology

The Metabolomics Platform is housed in approximately ~3,500 sq. ft. in the new facilities at 300 Binney Street, Cambridge, MA. We use high resolution orbitrap and triple quadrupole mass spectrometry systems in tandem with robust HPLC methods to generate data for targeted and nontargeted metabolomic studies. Current instrumentation includes:

  • High-Resolution, Accurate-Mass (HRAM) Orbitrap mass spectrometers (Thermo Q-Exactive Plus, Thermo Q-Exactive, Thermo HF, Thermo Focus, Thermo ID-X, Exploris 240)
  • Triple Quadrupole MS/MS systems (Sciex QTrap 7500, Sciex QTrap 5500, Agilent 6495)
  • HPLC/UPLC Liquid Chromatography (Agilent 1290 Infinty, Shimadzu Nexera 30, Waters Acquity Classic)
  • Agilent Single Quad MSD coupled to an Agilent Fraction Collector

Nontargeted LC-MS(/MS) Methods

Our team at the Broad has developed a suite of robust and precise liquid chromatography tandem mass spectrometry (LC–MS) methods that together profile over 1000 identified metabolites and lipids, as well as thousands of signals from unidentified compounds, in a single biospecimen. Below is our method shorthand name, column used, typical mass range, and target species.

  • HILIC-pos, Waters Atlantis Silica HILIC: (70-850 m/z) Polar metabolites detected in positive ion mode, including amines, amino acids, nucleosides, and other protonated or cationic metabolites
  • HILIC-neg, Phenomenex Luna NH2: (70-850 m/z) Polar metabolites detected in negative ion mode, including sugars, organic acids, purines, pyrimidines, and other anionic metabolites
  • C8-pos, Waters Acquity BEH C8: (220-1100 m/z) Lipids detected in positive ion mode, including phospholipids, sphingolipids, neutral lipids, and other hydrophobic lipid species
  • C18-neg, Waters Acquity HSS T3: (70-850 m/z) Intermediate-polarity to hydrophobic metabolites detected in negative ion mode, including free fatty acids, bile acids, oxidized lipids, and other anionic metabolites
  • C18-pos, Waters Acquity BEH C18: (70-850 m/z) Intermediate-polarity metabolites detected in positive ion mode

Acquisition is typically performed on Orbitrap instruments at 70,000 FWHM resolution with an approximately 8 Hz scan rate. Data are processed using both nontargeted feature extraction workflows, including Progenesis QI and a custom platform in development, for broad feature discovery, and targeted feature extraction using TFE to ensure that metabolites relevant to each study are consistently captured.

Targeted LC-MS Methods

In addition to Nontargeted LC-MS methods, we have developed customized targeted methods for the following biological pathways and chemical families:

  • Tryptophan pathway
  • Eicosanoids
  • Sterols

Samples are typically acquired using high sensitivity QQQ instruments, and features are confirmed by transition and comparison to MS1 RT standards.

Informatics and Processing

We apply a modular processing approach, enabled by our custom, cloud-based informatics platform we call Atlas. Atlas stores and indexes raw instrument files for projects, >5000 standards in our compound database, processed data, and provides our lab members an interactive web interface to visualize and manually curate data during processing. With the exception of feature extraction, all steps are are performed in Atlas, leading to 100%  tracebility and reproducibility.

  • Acquisition - During acquisition, Automated and Manual DailyQC are performed to monitor acquisition queues.
  • Backup - Raw instrument files uploaded backed up into two distinct cloud locations, our disaster recovery Cold Storage and Atlas.
  • Nontargeted Feature Extraction - Nontargeted feature extraction is perform, producing raw feature tables which are uploaded to Atlas.
  • Annotation - Features are annotated via authenticated MS1 and MS2 standards and manually curated.
  • Targeted Extraction - Targeted extraction is performed to extract difficult-to-integrate features using TFE.
  • Drift Correction - Data is drift corrected using pooled reference technical replicates.
  • Filtering - Finally, on demand, data are filtered, clustered, and formatted for sharing.
Lab Space