Abstract: Described is a method for diagnosing a person having or being at risk of developing certain B-cell pathologies, including Sjögren's Syndrome and non-Hodgkin's lymphoma, and excluding patients with symptoms similar to Sjögren's Syndrome but with a different etiology, comprising the following steps: providing a sample of a body fluid or tissue from said person, said sample containing a mixture of unknown proteins, protein fragments or peptides; analyzing said samples with mass spectrometry to generate a m/z (mass to charge ratio) spectrogram for each sample; comparing whether the patient's sample contains m/z values that are characteristic of a Sjögren's Syndrome reference database derived from the analysis and cataloguing of multiple patient spectrograms; and determining whether said patient either has or does not have Sjögren's Syndrome on the basis of this comparative analysis.

Abstract: Described is a method for diagnosing a person having or being at risk of developing Sjögren's Syndrome and excluding patients with symptoms similar to Sjögren's Syndrome but with a different etiology, comprising the following steps: providing a sample of a body fluid or tissue from said person, said sample containing a mixture of unknown proteins, protein fragments or peptides; analyzing said samples with mass spectrometry to generate a m/z (mass to charge ratio) spectrogram for each sample; comparing whether the patient's sample contains m/z values that are characteristic of a Sjögren's Syndrome reference database derived from the analysis and cataloguing of multiple patient spectrograms; and determining whether said patient either has or does not have Sjögren's Syndrome on the basis of this comparative analysis.

Abstract: Electronic data is classified using adaptive focused replicator networks (AFRNs). AFRNs are sets of array elements, each array element being trainable in order to replicate a predetermined sub-group of data. Unknown data is inputted into each and every array element in an AFRN array and then replicated by each element array. A comparison is made for each array element to determine the accuracy of each replication. If only one array element successfully replicates the unknown data, then the unknown data is classified in accordance with the corresponding predetermined sub-group of data.