ABSTRACT More than 3 million tests are performed each year to support the laboratory diagnosis of human Lyme disease (LD). While the CDC conventional standard two-tier (CSTT) approach for serodiagnosis of LD has worked relatively well when used as recommended, there is plenty of room for improvement. Of a number of weaknesses associated with the supplemental immunoblot of the CSTT the most significant is low reproducibility due to the subjective visual interpretation of results. To overcome these weaknesses the CDC recently updated its recommendations based on a modified STT (MSTT) in that a second EIA can replace the immunoblot. The major goal of this project is to develop an objective, quantitative, multiplex EIA that can detect four antibody isotypes (IgM/D/G/A) and all four IgG subclasses (IgG1/2/3/4) to leverage acquisition of simultaneous antibody profile information on multiple B. burgdorferi antigens to build an assay that can discriminate Lyme disease stage with increased overall sensitivity without incurring in loss of specificity. The novelty of this study relies on: 1) evaluation of B. burgdorferi antigen-specific antibody isotypes and IgG subclasses that can be correlated with Lyme disease stage; and 2) development of new diagnostic tools using machine learning techniques to train and integrate all data and produce an objective result to discriminate early Lyme from early disseminated/late Lyme disease. We expect this Phase I SBIR to allow us to develop a new EIA for serodiagnosis of Lyme disease (isoEIAplex-Ld) and to further an ongoing collaboration with DCN diagnostics for the adaptation of our biomarkers to a new rapid Lateral Flow Assay (see Letter of Support) for a follow up Phase II SBIR .