Methods and compositions for analyte quantification

A method for enabling high-throughput quantitation of biomolecules, comprising:

• (a) combining one or more calibrators with a mixed sample comprising one or more target analytes;
• (b) separating the mixed sample using a chromatographic column;
• (c) obtaining raw mass spectrometry (MS) data of the separated sample of from (b);
• and
• (d) processing the raw MS data using a machine learning model means to directly determine, in less than 12 minutes of total runtime, an absolute concentration of the one or more target analytes in the mixed sample;
• wherein the machine learning model is trained using a training dataset comprising raw, unprocessed MS training data further comprising:
• (i) a first set of intensity values for one or more reference analytes having a known concentration and (ii) a second set of intensity values for one or more reference calibrators having a known concentration;
• wherein the raw MS data for the mixed sample comprises a first time series MS signal for the one or more target analytes and a second time series MS signal for the one or more calibrators;
• and
• wherein the machine learning model processes the raw MS data to determine an absolute concentration of the one or more target analytes, based at least in part on a relationship or correlation between the first time series MS signal and the second time series MS signal.