Alternative splicing (AS) contributes to the functional diversity of the genome by allowing multiple transcript variants to be encoded by one gene. Prior work using transcriptomics has shown that AS is highly tissue specific and has strong potential to alter protein interaction networks. In contrast, knowledge into the protein-level existence and function of the translated protein isoforms remains lagging. To address this gap, we and others have employed a proteogenomics approach that combines information from transcriptomics and proteomics to identify hundreds of AS-derived isoforms in mammalian tissues and cell lines. We have used our in-house software JCAST to translate RNA-seq-derived splicing information into protein isoform sequence databases and then generated new mass spectrometry data to discover AS-derived isoforms, leading to the discovery of protein-level evidence of hundreds of AS-derived protein isoforms. More recently, we found 29 non-canonical isoforms that show signs of regulation based on statistically significant preferences in different chambers of the human heart from public proteomics datasets. To investigate the potential function of these non-canonical isoforms, we focused on the atrium-enriched isoform 2 of PDZ and LIM Domain 3 (PDLIM3-2). Using recent deep learning based software tools including AlphaFold-Multimer, we examined the function of the expressed protein isoforms and found that the alternative region of the protein impacts PDLIM3 binding to its interaction partner. Moreover, on a proteome-wide scale, we find that variant regions that differ between alternative and canonical isoforms are highly enriched in intrinsically disordered regions (IDR), and over two-thirds of such regions are predicted to function in protein binding and/or RNA binding. Our analysis lends further credence to the notion that AS diversifies the proteome by rewiring IDRs, which are increasingly recognized to play important roles in the generation of biological function from protein sequences.