Lee, N. F. | bioRxiv Science

bioRxiv2026

Antibody Profiles in Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections

Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections (PANDAS) is characterized by prepubertal abrupt onset of obsessive-compulsive disorder (OCD). The sine qua non is group A streptococcus (GAS) infection, which is hypothesized to elicit an IgG-class anti-GAS antibody response that cross-reacts with antigens in the basal ganglia. However, the association between GAS antibody (GAS-IgG) levels and PANDAS has been inconsistent, and qualitative differences in GAS-IgG profiles have not been carefully evaluated in well-phenotyped cohorts. Moreover, independent studies have yet to converge on anti-neural autoantibodies that are specific to PANDAS. Here, we used phage display immunoprecipitation sequencing (PhIP-Seq) to perform ultra-deep anti-pathogen antibody repertoire profiling of serum from definitive pediatric PANDAS patients (N = 34) collected as part of a prior double-blind, placebo-controlled clinical trial of intravenous immunoglobulin (IVIg). PANDAS cases were compared to pediatric controls without a history of neuropsychiatric illness (N = 31). To assess for objective evidence of neuroglial injury, serum neurofilament light (NfL) and glial fibrillary acidic protein (GFAP) levels were compared to healthy pediatric controls. Within PANDAS, NfL and GFAP levels were compared between pre- and post-treatment sera. To evaluate for central autoantibodies, a subset of baseline cerebrospinal fluid (CSF) samples (N = 25) was profiled by full-length human protein microarray. Though GAS reactivity by PhIP-Seq was well correlated with clinical anti-DNaseB and anti-streptolysin O titers, there were no quantitative or qualitative differences in GAS-IgG profiles between PANDAS and controls. Furthermore, NfL and GFAP levels did not differ between cases and controls. Within PANDAS, changes in NfL or GFAP levels at six weeks did not differ between placebo and IVIg groups. However, CSF autoantibody profiling by protein microarray revealed infrequent but notable candidate autoantibodies. In one patient, we identified autoantibodies against Argonaute family proteins (AGO-IgG), a marker of autoimmune sensory neuropathy. Longitudinal measurement of AGO-IgG in sera revealed that titers were unchanged after placebo, but decreased after IVIg, coinciding with symptomatic improvement, including a decrease in that patients CY-BOCS score. Overall, these results do not support an etiologic role for GAS-IgG in PANDAS. However, some individuals diagnosed with PANDAS may harbor anti-neural autoantibodies.

immunology↗

bioRxiv2025

Machine learning driven acceleration of biopharmaceutical formulation development using Excipient Prediction Software (ExPreSo)

Formulation development of protein biopharmaceuticals has become increasingly challenging due to new modalities and higher target drug substance concentrations. The limited amount of drug substance available during development, coupled with extensive analytical requirements, restrict the number of excipients that can be empirically screened. There is a strong need for in silico tools to optimize excipient pre-selection before wet lab experiments. Here, we introduce Excipient Prediction Software (ExPreSo), a supervised machine learning algorithm that suggests excipients based on the properties of the protein drug substance and target product profile. ExPreSo was trained on a dataset comprising 335 regulatory-approved peptide and protein drug products. Predictive features included protein structural properties, protein language model embeddings, and drug product characteristics. ExPreSo showed good performance for the nine most prevalent excipients in biopharmaceutical formulations and minimal overfitting. A fast variant of ExPreSo using only sequence-based input features showed similar prediction power to slower models that relied on molecular modeling. Notably, an ExPreSo variant using only protein-based input features also showed good performance, indicating resilience to the influence of platform formulations. To our knowledge, this is the first machine learning algorithm to suggest biopharmaceutical excipients based on the dataset of regulatory-approved drug products. Overall, ExPreSo shows great potential to reduce the time, costs, and risks associated with excipient screening during formulation development. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=77 SRC="FIGDIR/small/637685v3_ufig1.gif" ALT="Figure 1"> View larger version (29K): org.highwire.dtl.DTLVardef@2e3e5corg.highwire.dtl.DTLVardef@342aorg.highwire.dtl.DTLVardef@160df8forg.highwire.dtl.DTLVardef@f547d8_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗