Abstract: Methods and apparatus are provided for generating an embedding of a graph. The graph includes a plurality of nodes and each node includes a connection to another one or more of the nodes. The method including and/or apparatus configured to: receiving data representative of at least a portion of the graph; transforming the nodes of the graph into a non-Euclidean geometry; iteratively updating an embedding model based the transformed nodes in the non-Euclidean geometry based on a causal loss function and a link prediction function associated with the non-Euclidean geometry.

Abstract: A computer-implemented method of stratifying a population of patients into disease endotypes is provided. The method comprises: encoding data relating to the patients as latent variables; determining one or more importance measures of the latent variables; prioritising the latent variables using the importance measures; interpreting one or more of the ranked latent variables; and identifying a disease endotype that is represented by one or more of the interpreted latent variables.

Abstract: A computer-implemented method and a system of selecting a cell line for an assay. The computer-implemented method and system encode data, which is comprised of one or more features, as one or more latent variables. The one or more features encoded in the one or more latent variables are identified and mapped to cell lines based on the one or more features. A relevance of one or more targets to each of one or more of the one or more latent variables is determined and the one or more targets to the cell lines are matched via the one or more latent variables.

Abstract: An optoelectronic device includes a semiconductor die that includes a substrate layer, a laser diode, first and second conducting pads, a cathode pad, an anode pad, and a passivation layer. The laser diode and the conducting pads are formed on the substrate layer. The formation of the conducting pads directly on the substrate layer offers an increased area for heat dissipation. The cathode pad is formed on the first conducting pad whereas the anode pad is formed above the second conducting pad. The passivation layer is formed above the laser diode. The attachment of the semiconductor die to a submount of the optoelectronic device occurs by way of the cathode pad and the anode pad. After the attachment, a free space is created directly between the passivation layer and the submount to reduce the impact of solder bonding stress on the laser diode.

Abstract: A photonic integrated circuit (PIC) includes various mode field adapters (MFAs), a waveguide, and various contact pads. All the MFAs are on a same facet of the PIC. One MFA of the PIC outputs a first optical signal that is an amplified version of a second optical signal. The waveguide is divided into two waveguide arms and a bend portion to join the two waveguide arms. The waveguide extends between the MFAs such that the second optical signal propagates through the waveguide. Further, each waveguide arm is formed between the contact pads. The second optical signal propagating through the waveguide is amplified based on a current that is injected in the PIC by way of the contact pads.

Abstract: Methods and apparatus are provided for generating a graph neural network (GNN) model based on an entity-entity graph. The entity-entity graph comprising a plurality of entity nodes in which each entity node is connected to one or more entity nodes of the plurality of entity nodes by one or more corresponding relationship edges. The method comprising: generating an embedding based on data representative of the entity-entity graph for the GNN model, wherein the embedding comprises an attention weight assigned to each relationship edge of the entity-entity graph; and updating weights of the GNN model including the attention weights by minimising a loss function associated with at least the embedding; wherein the attention weights indicate the relevancy of each relationship edge between entity nodes of the entity-entity graph. The entity-entity graph may be filtered based on the attention weights of a trained GNN model.