In the last few years, machine learning (ML) and artificial intelligence have seen a new wave of publicity fueled by the huge and ever‐increasing amount of data and computational power as well as the discovery of improved learning algorithms. However, the idea of a computer learning some abstract concept from data and applying them to yet unseen situations is not new and has been around at least since the 1950s. Many of these basic principles are very familiar to the pharmacometrics and clinical pharmacology community. In this paper, we want to introduce the foundational ideas of ML to this community such that readers obtain the essential tools they need to understand publications on the topic. Although we will not go into the very details and theoretical background, we aim to point readers to relevant literature and put applications of ML in molecular biology as well as the fields of pharmacometrics and clinical pharmacology into perspective.

Abstract. Exosomes are important contributors to cellMetastases are responsible for the death of a large majority of cancer patients despite considerable progress in surgical techniques, radiotherapy, chemotherapy and targeted therapies including immuno-therapy (1). A dramatic reduction of metastatic burden has been observed, however, tumor elimination is almost always incomplete. This phenomenon is based on drug resistance, which is due to adaptation of intracellular pathways or on activation of survival-supporting autocrine and paracrine pathways and several secreted factors expressed by drug-sensitive tumor cells after therapy (2). Metastasis can occur through release of cancer cells from the primary tumor into body cavities that holds true for ovarian and CNS tumors, or via hematogeneous and lymphatic vessels of the circulatory system (3). Metastases of some tumors are directed besides to lymph nodes to mainly one type of organ only, such as prostate cancer to the bones, pancreatic cancer and uveal melanoma to the liver, whereas tumors such as melanoma, breast-and lung cancer can colonize several types of organs (3). Tumor cells have been found in the blood of patients with early-stage cancer and in some cases even before the primary tumor has been diagnosed (4, 5). Recently, making use of single-cell expression profiling, it has been shown that early metastatic cells possess a stem-like gene signature and give rise to heterogeneous metastases (6). The metastatic process is characterized by a defined sequence of events (7,8). Initial steps are detachment from the extracellular matrix (ECM), invasion into the surrounding tissue and proteolysis of the basement membrane. After intravasation, survival of circulating tumor cells (CTCs) is achieved by forming clusters, binding to platelets and immune evasion. Subsequently, they arrest in distal microvascular beds and extravasation can be achieved either by migration through intercellular junctions of endothelial cells (EC) or penetration of a single EC (9). CTCs can also colonize their tumors of origin, a process referred to as "tumor self-seeding", selecting for cancer cell populations more aggressive than those present in the primary tumor (10). After extravasation, colonization and outgrowth in the parenchyma of distant organs are the following steps. A common theme of the metastatic process is settlement of disseminated tumor cells (DTCs) into latency (dormancy), which can last from several months to decades (11). It has been observed that DTCs are recruited into pre-metastatic niches that support their survival by interactions with endothelial, myeloid cells, fibroblasts and other types of cells. After adaptation to the host microenvironment and suppression of an anti-tumoral immune response, DTCs are activated by not yet completely resolved mechanisms. Finally their outgrowth is based on an angiogenic switch mediated by pro-angiogenic factors and establishment of a vascular network to support the metabolic demands of colonizing tumor cells (12). In the follo...

Fully differentiated pancreatic β cells are essential for normal glucose homeostasis in mammals. Dedifferentiation of these cells has been suggested to occur in type 2 diabetes, impairing insulin production. Since chronic fuel excess (“glucotoxicity”) is implicated in this process, we sought here to identify the potential roles in β-cell identity of the tumor suppressor liver kinase B1 (LKB1/STK11) and the downstream fuel-sensitive kinase, AMP-activated protein kinase (AMPK). Highly β-cell-restricted deletion of each kinase in mice, using an Ins1-controlled Cre, was therefore followed by physiological, morphometric, and massive parallel sequencing analysis. Loss of LKB1 strikingly (2.0–12-fold, E<0.01) increased the expression of subsets of hepatic (Alb, Iyd, Elovl2) and neuronal (Nptx2, Dlgap2, Cartpt, Pdyn) genes, enhancing glutamate signaling. These changes were partially recapitulated by the loss of AMPK, which also up-regulated β-cell “disallowed” genes (Slc16a1, Ldha, Mgst1, Pdgfra) 1.8- to 3.4-fold (E<0.01). Correspondingly, targeted promoters were enriched for neuronal (Zfp206; P=1.3×10−33) and hypoxia-regulated (HIF1; P=2.5×10−16) transcription factors. In summary, LKB1 and AMPK, through only partly overlapping mechanisms, maintain β-cell identity by suppressing alternate pathways leading to neuronal, hepatic, and other characteristics. Selective targeting of these enzymes may provide a new approach to maintaining β-cell function in some forms of diabetes.—Kone, M., Pullen, T. J., Sun, G., Ibberson, M., Martinez-Sanchez, A., Sayers, S., Nguyen-Tu, M.-S., Kantor, C., Swisa, A., Dor, Y., Gorman, T., Ferrer, J., Thorens, B., Reimann, F., Gribble, F., McGinty, J. A., Chen, L., French, P. M., Birzele, F., Hildebrandt, T., Uphues, I., Rutter, G. A. LKB1 and AMPK differentially regulate pancreatic β-cell identity.

Abstract.Metastasis is the major cause of death in patients with cancer. It is mediated by a multi-step process referred to as the metastatic cascade (1, 2). Initial steps include local invasion and migration, angiogenesis, epithelialmesenchymal transition (EMT) and intravasation. Tumor cells enter the circulation as single cells or circulating tumor cell clusters, are coated by platelets to escape an immune response and subsequently arrest in capillaries in distant organs as a prerequisite for extravasation. Colonization starts by homing of tumor cells in supporting niches of the organ parenchyma, followed by a latency phase which can last from several months to decades. A prerequiste for overt outgrowth of micrometastases is their adaptation to the local microenvironment and acquisition of colonisation-promoting traits (3-6). The pattern of colonized distant organs depends on the tumor type and can range from predominant spread to one organ and colonization of different types of organs sequentially or simultaneously (7). Several genes and their products have been identified to mediate crucial steps of the metastatic process such as metastasis initiation and progression, as well as organ-specific functions of metastasis (virulence) (8, 9). These gene products include proteases, chemokines, cytokines and their receptors, angiogenic factors, intracellular and transmembrane kinases, adhesion molecules, components of the extracellular matrix (ECM), GPI-linked receptors and carbohydrate metabolism-related enzymes (8, 9). More recently, an important impact of RNA-related molecules for metastasis has emerged. MicroRNAs (miRs) and other types of RNA modulate metastasis via regulatory networks (10,11). In this review we focus on the role of long non-coding RNAs (lncRNA) as promoters or inhibitors of metastasis in different tumor entities since there is an urgent need to define new targets for therapeutic intervention. With the exception of denosumab for treatment of bone metastases, all other agents evaluated in clinical studies for treatment of metastatic disease gave rise to mixed or disappointing results (6).

Abstract. Ovarian