BMXP - The Metabolomics Platform at the Broad Institute

pip install bmxp

Please cite: https://www.biorxiv.org/content/10.1101/2023.06.09.544417v1.full

This is a collection of tools for processing our data, which powers our cloud processing workflow. Each tool is meant to be a standalone module that performs a step in our processing pipeline. They are written in Python and C, and designed to be perfomant and cloud-compatible.

Eclipse - Align two or more same-method nontargeted LCMS datasets
Gravity - Cluster redundant LCMS features based on RT and Correlation (And someday, XIC shape)
Blueshift - Drift Correction via pooled technical replicates and internal standards
Formation - Formatting and Final QC
Chroma - Read .raw and .mzml files

We expect users to be familiar with Python and already have an understanding of LCMS Metabolomics data processing and the specific steps they wish to accomplish.

While the tools are and always will be standalone, we are working on linking them closer together with a shared schema, and eventually may have a pipeline ability to run all steps, given a set of parameters.

We are open to feedback and suggestions, with a focus on performance and application in pipelines.

Shared Schema

All BMXP modules use a shared schema and file formats with our prefered columns headers. These files are (along with their labels):

Feature Metadata bmxp.FMDATA - Describes the feature. Index default is Compound_ID
Injection Metadata bmxp.IMDATA - Describes the Injection. Index default is injection_id
Sample Metadata bmxp.SMDATA - Describes the biospecimen from which the Injection is derived. Index default is broad_id
Feature Abundances - Pivot table of Feature x Injection (Compound_ID x injection_id) containing the abundances.

Some modules (Blueshift, Eclipse) require merging Feature Metadata + Feature Abundances.

These can be changed globally so that all packages will use the same terminology. To update the schema, modify the dictionary objects in the module directly prior to running code. For example:

import bmxp
from bxmp.eclipse import MSAligner
from bxmp.blueshift import DriftCorrection
from bmxp.gravity import cluster
bmxp.FMDATA['Compound_ID'] = 'Feature_ID'
bmxp.IMDATA['injection_id'] = 'Filename'

# continue with work...

With those changes above, Eclipse, Blushift and Gravity will use "Feature_ID" and "Filename" as column headers instead of "Compound_ID" and "injection_id".

Feature Metadata - bmxp.FMDATA

Feature Metadata describes the LCMS feature. This is a mixture of fundamental nontargeted feature information, annotation info, and anything else.

Feature Specific

Compound_ID - Index, Project-unique feature ID (a bit of a misnomer)
RT - Unitless retention time, may or may not be scaled
MZ - Unsigned mass-to-charge ratio
Intensity - Average feature intensity
Method - Human Readable name of LCMS method used
__extraction_method - Name of extraction method/software used. Used to denote mixed Targeted/Nontargeted

Annotation

Annotation_ID - Method-unique annotation label
Adduct - Adduct form of the annotation
__annotation_id - Globally unique annotation identifier
Metabolite - Preferred display/reporting name of metabolite
Non_Quant - Boolean denoting that a feature is not quanitifiable

Generated by Gravity

Cluster_Num - Cluster number assigned during Gravity clustering
Cluster_Size - Number of members in the cluster

Generated by Blueshift

Batches Skipped - Batches that were skipped due to lack of PREFs

Injection Metadata - bmxp.IMDATA

injection_id - Index, Injection name, usually filename without the extension
broad_id - Assigned biospeciemn label
program_id - Biospecimen label as received (inherited from Sample Metadata)
injection_type - Type of injection ("sample", "prefa", "prefb", "blank", "other-", "not_used-")
comments - Comments about the injection
column_number - Column number, in multi-column studies
injection_order - Injection number, not skipping blanks or non-samples
batches - Denotes batches ('batch_start' or 'batch_end')

Generated by Blueshift

QCRole - Role in drift correction ("QC-drift_correction", "QC-pooled_ref", "QC-not_used", "sample")

Sample Metadata - bmxp.SMDATA

broad_id - Assigned biospecimen label
Arbitrary Metadata Columns - Any column label except labels in Injection Metadata

Metadata

Release files for bmxp 0.5.6

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for bmxp 0.5.6
File	Size	Uploaded
bmxp-0.5.6.tar.gz	2.7 MB	Sep 18, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for bmxp 0.5.6
File	Interpreter	ABI	Platform	Reset
bmxp-0.5.6-py3-none-any.whl 1.2 MB Sep 18, 2026	Python 3	none	any	Details

Total release size: 3.9 MB

Release files / bmxp-0.5.6.tar.gz

Download URL	bmxp-0.5.6.tar.gz
Size	2.7 MB
Tags	Source
SHA-256 checksum How to use checksums	`59fcb52e4c7c15189a8946c265eb8340e411cfdf9ac72a28c97fe51fd8e35ca5`
BLAKE2b-256 checksum How to use checksums	`f58b98e431dcf84343589b2cc8917e86c4060d3fa003a9bf10676b4c9a53da5a`
Upload date	Sep 18, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.13`

Release files / bmxp-0.5.6-py3-none-any.whl

Download URL	bmxp-0.5.6-py3-none-any.whl
Size	1.2 MB
Tags	Python 3
SHA-256 checksum How to use checksums	`0c2d5895010736f9af89e182d4c76aa85dbb1e69a853890da18ce3e7890c70a4`
BLAKE2b-256 checksum How to use checksums	`234d74d0a12b2e40339b5d29d1066292296771671016ec507295482d68a9c55c`
Upload date	Sep 18, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.13`