Tumor-specific mutations form novel immunogenic peptides called neoantigens. Neoantigens can be used as a biomarker predicting patient response to cancer immunotherapy. Although a predicted binding affinity (IC₅₀) between peptide and major histocompatibility complex class I is currently used for neoantigen prediction, large number of false-positives exist. We developed Neopepsee, a machine-learning-based neoantigen prediction program for next-generation sequencing data. With raw RNA-seq data and a list of somatic mutations, Neopepsee automatically extracts mutated peptide sequences and gene expression levels. We tested 14 immunogenicity features to construct a machine-learning classifier and compared with the conventional methods based on IC₅₀ regarding sensitivity and specificity. We tested Neopepsee on independent datasets from melanoma, leukemia, and stomach cancer. Nine of the 14 immunogenicity features that are informative and inter-independent were used to construct the machine-learning classifiers. Neopepsee provides a rich annotation of candidate peptides with 87 immunogenicity-related values, including IC₅₀, expression levels of neopeptides and immune regulatory genes (e.g. PD1, PD-L1), matched epitope sequences, and a three-level (high, medium, and low) call for neoantigen probability. Compared with the conventional methods, the performance was improved in sensitivity and especially two- to threefold in the specificity. Tests with validated datasets and independently proven neoantigens confirmed the improved performance in melanoma and chronic lymphocytic leukemia. Additionally, we found sequence similarity in proteins to known pathogenic epitopes to be a novel feature in classification. Application of Neopepsee to 224 public stomach adenocarcinoma datasets predicted ∼7 neoantigens per patient, the burden of which was correlated with patient prognosis. Neopepsee can detect neoantigen candidates with less false positives and be used to determine the prognosis of the patient. We expect that retrieval of neoantigen sequences with Neopepsee will help advance research on next-generation cancer immunotherapies, predictive biomarkers, and personalized cancer vaccines.

The role of brain somatic mutations in Alzheimer’s disease (AD) is not well understood. Here, we perform deep whole-exome sequencing (average read depth 584×) in 111 postmortem hippocampal formation and matched blood samples from 52 patients with AD and 11 individuals not affected by AD. The number of somatic single nucleotide variations (SNVs) in AD brain specimens increases significantly with aging, and the rate of mutation accumulation in the brain is 4.8-fold slower than that in AD blood. The putatively pathogenic brain somatic mutations identified in 26.9% (14 of 52) of AD individuals are enriched in PI3K-AKT, MAPK, and AMPK pathway genes known to contribute to hyperphosphorylation of tau. We show that a pathogenic brain somatic mutation in PIN1 leads to a loss-of-function mutation. In vitro mimicking of haploinsufficiency of PIN1 aberrantly increases tau phosphorylation and aggregation. This study provides new insights into the genetic architecture underlying the pathogenesis of AD.

In human FMCD tissue, a small fraction of pS6+ neurons are enriched for somatic activating mutations of the PI3K-AKT-mTOR pathway. Sparse electroporation of the AKT3 mutation into the developing mouse brain causes a reelin-dependent, non–cell autonomous disruption of neuronal migration, leading to impaired cortical lamination and seizure-like epileptiform EEG activity. Focal malformations of cortical development (FMCDs) account for the majority of drug-resistant pediatric epilepsy. Postzygotic somatic mutations activating the phosphatidylinositol-4,5-bisphosphate-3-kinase (PI3K)–protein kinase B (AKT)–mammalian target of rapamycin (mTOR) pathway are found in a wide range of brain diseases, including FMCDs. It remains unclear how a mutation in a small fraction of cells disrupts the architecture of the entire hemisphere. Within human FMCD-affected brain, we found that cells showing activation of the PI3K-AKT-mTOR pathway were enriched for the AKT3E17K mutation. Introducing the FMCD-causing mutation into mouse brain resulted in electrographic seizures and impaired hemispheric architecture. Mutation-expressing neural progenitors showed misexpression of reelin, which led to a non–cell autonomous migration defect in neighboring cells, due at least in part to derepression of reelin transcription in a manner dependent on the forkhead box (FOX) transcription factor FOXG1. Treatments aimed at either blocking downstream AKT signaling or inactivating reelin restored migration. These findings suggest a central AKT-FOXG1-reelin signaling pathway in FMCD and support pathway inhibitors as potential treatments or therapies for some forms of focal epilepsy.