Abstract: Techniques for generating smart replies involving image files are disclosed herein. In some example embodiments, a computer system detects that a first message comprising a first image file has been transmitted from a first computing device of a first user to a second computing device of a second user, and generates a first plurality of smart replies based on a first embedding vector of the first image file, where the first embedding vector of the first image file is based on at least one of first textual metadata of the first image file, first image data of the first image file, and a first set of query text used by a first set of users in a first set of searches that resulted in the first image file being included in a first set of transmitted messages.

Abstract: A distributed, online machine learning system is presented. Contemplated systems include many private data servers, each having local private data. Researchers can request that relevant private data servers train implementations of machine learning algorithms on their local private data without requiring de-identification of the private data or without exposing the private data to unauthorized computing systems. The private data servers also generate synthetic or proxy data according to the data distributions of the actual data. The servers then use the proxy data to train proxy models. When the proxy models are sufficiently similar to the trained actual models, the proxy data, proxy model parameters, or other learned knowledge can be transmitted to one or more non-private computing devices. The learned knowledge from many private data servers can then be aggregated into one or more trained global models without exposing private data.

Abstract: Contemplated systems and methods allow for prediction of chemotherapy outcome for patients diagnosed with high-grade bladder cancer. In particularly preferred aspects, the prediction is performed using a model based on machine learning wherein the model has a minimum predetermined accuracy gain and wherein a thusly identified model provides the identity and weight factors for omics data used in the outcome prediction.

Abstract: Techniques for starting an electronic communication based on a captured image are disclosed herein. In some embodiments, a computer system detects that an image has been captured by a camera on a first mobile device of a first user, where the captured image has been captured by the camera at a point in time, and, in response to detecting that the image has been captured, the computer system identifies at least one other user in the captured image. In some example embodiments, the computer system transmits a message to an electronic destination associated with the other user(s) based on the identifying of the other user(s) in the captured image.

Abstract: The present inventive subject matter provides apparatus, systems, and methods in which a diagnostic test is identified, where the diagnostic test is for determining whether a particular treatment is effective for a particular patient based on one or more characteristics of a patient's cells. When a treatment is developed with the potential to treat one or more diseases, the drug can have different effects on different cell lines related to the diseases. A machine learning system is programmed to infer a measurable cell characteristic, out of many different measurable cell characteristics, that has a desirable correlation with the sensitivity data of different cell lines to a treatment. The machine learning system is programmed to then determine, based on the correlation, a threshold level of the cell characteristic the patient should exhibit in order to recommend administering the treatment.

Abstract: Transcriptomics data from tumor tissue of patients diagnosed with metastatic breast cancer are clustered and associated with overall survival of the patients. A subset of genes from one of the cluster associated with poor outcome are used to generate a survival prediction model predicting a survival time based on expression levels of a plurality of genes. Using such generated survival prediction model, a survival time of a patient diagnosed with metastatic breast cancer can be predicted and a treatment regimen can be updated or generated based on the survival time.

Abstract: Described herein is a technique to facilitate the sharing of a content item presented in a content feed of a social networking service. Upon detecting that a member has selected an option to share a content item, a content sharing interface is presented. The content sharing interface includes options share the content item publically via a content feed, and privately via a messaging service. The content sharing interface provides a ranked list of recommended recipients, where the recommended recipients in the list are selected and ordered based on several factors, including factors relating to the relationship between the sharing member and the recommended recipients, as well as factors relating to the subject matter of the content item and the likelihood that a recommended recipient would be interested in the content item.

Abstract: Methods are provided for identifying whether a cancer patient, and especially a breast cancer patient, will be responsive to treatment. Specified TOPO2A, IDO1 and/or p16 fragment peptides are precisely detected and quantitated by SRM-mass spectrometry directly in cancer cells collected from tumor tissue that was obtained from a cancer patient and compared to reference levels in order to determine if the cancer patient will positively respond to treatment. Measurement of TOPO2A provides a direct indication of whether a patient will respond to anthracycline-containing therapy, and, in particular, neoadjuvant anthracycline-containing therapy. Quantitative levels of IDO1 and p16 are compared to reference levels in order to determine if a breast cancer patient will likely demonstrate a pathologically complete response (pCR) of cancer after cancer therapy treatment, irrespective of the chosen treatment.

Abstract: Improved accuracy of SNV-based genetic tests is performed using DNA sequencing data from a tumor sample and a matched normal sample to determine SNVs, and RNA sequencing data from the tumor sample are used to ascertain expression of so identified SNVs.

Abstract: Techniques are provided for predicting DNA accessibility. DNase-seq data files and RNA-seq data files for a plurality of cell types are paired by assigning DNase-seq data files to RNA-seq data files that are at least within a same biotype. A neural network is configured to be trained using batches of the paired data files, where configuring the neural network comprises configuring convolutional layers to process a first input comprising DNA sequence data from a paired data file to generate a convolved output, and fully connected layers following the convolutional layers to concatenate the convolved output with a second input comprising gene expression levels derived from RNA-seq data from the paired data file and process the concatenation to generate a DNA accessibility prediction output. The trained neural network is used to predict DNA accessibility in a genomic sample input comprising RNA-seq data and whole genome sequencing for a new cell type.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for batching requests for presence information. A client device determines a set of user indicators that are displayed on a display of the client device during a period of time. Each user indicator corresponds to a user of an online service, and is presented on the display as a result of a user of the client device viewing content provided by the online service. At completion of the period of time, the client device determines a set of remaining user indicators displayed on the display of the client device, and transmits, to a server associated with the online service, a batched request for presence information for a set of users corresponding to the set of remaining user indicators. The client device presents presence indicators indicating the status of the users based on the requested presence information.

Abstract: Molecular profiles of metastatic tumors can be more accurately determined using combination of DNA mutational profiles and RNA expression profiles of selected genes that show substantial changes upon anti-tumor treatment. Such combinatorial information can be used to determine differential pathway activity that may be related to the sensitivity or resistance to the anti-tumor treatment.

Abstract: Techniques for starting an electronic communication based on a captured image are disclosed herein. In some embodiments, a computer system detects that an image has been captured by a camera on a first mobile device of a first user, where the captured image has been captured by the camera at a point in time, and, in response to detecting that the image has been captured, the computer system identifies at least one other user in the captured image. In some example embodiments, the computer system transmits a message to an electronic destination associated with the other user(s) based on the identifying of the other user(s) in the captured image.

Abstract: Techniques for generating smart replies involving image files are disclosed herein. In some example embodiments, a computer system detects that a first message comprising a first image file has been transmitted from a first computing device of a first user to a second computing device of a second user, and generates a first plurality of smart replies based on a first embedding vector of the first image file, where the first embedding vector of the first image file is based on at least one of first textual metadata of the first image file, first image data of the first image file, and a first set of query text used by a first set of users in a first set of searches that resulted in the first image file being included in a first set of transmitted messages.

Abstract: Techniques are provided for predicting DNA accessibility. DNase-seq data files and RNA-seq data files for a plurality of cell types are paired by assigning DNase-seq data files to RNA-seq data files that are at least within a same biotype. A neural network is configured to be trained using batches of the paired data files, where configuring the neural network comprises configuring convolutional layers to process a first input comprising DNA sequence data from a paired data file to generate a convolved output, and fully connected layers following the convolutional layers to concatenate the convolved output with a second input comprising gene expression levels derived from RNA-seq data from the paired data file and process the concatenation to generate a DNA accessibility prediction output. The trained neural network is used to predict DNA accessibility in a genomic sample input comprising RNA-seq data and whole genome sequencing for a new cell type.

Abstract: Techniques for starting an electronic communication based on a captured image are disclosed herein. In some embodiments, a computer system detects that an image has been captured by a camera on a first mobile device of a first user, where the captured image has been captured by the camera at a point in time, and, in response to detecting that the image has been captured, the computer system identifies at least one other user in the captured image. In some example embodiments, the computer system transmits a message to an electronic destination associated with the other user(s) based on the identifying of the other user(s) in the captured image.

Abstract: Techniques for starting an electronic communication based on a captured image are disclosed herein. In some embodiments, a computer system detects that an image has been captured by a camera on a first mobile device of a first user, where the captured image has been captured by the camera at a point in time, and, in response to detecting that the image has been captured, the computer system identifies at least one other user in the captured image. In some example embodiments, the computer system transmits a message to an electronic destination associated with the other user(s) based on the identifying of the other user(s) in the captured image.

Abstract: Techniques are provided for predicting DNA accessibility. DNase-seq data files and RNA-seq data files for a plurality of cell types are paired by assigning DNase-seq data files to RNA-seq data files that are at least within a same biotype. A neural network is configured to be trained using batches of the paired data files, where configuring the neural network comprises configuring convolutional layers to process a first input comprising DNA sequence data from a paired data file to generate a convolved output, and fully connected layers following the convolutional layers to concatenate the convolved output with a second input comprising gene expression levels derived from RNA-seq data from the paired data file and process the concatenation to generate a DNA accessibility prediction output. The trained neural network is used to predict DNA accessibility in a genomic sample input comprising RNA-seq data and whole genome sequencing for a new cell type.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for batching requests for presence information. A client device determines a set of user indicators that are displayed on a display of the client device during a period of time. Each user indicator corresponds to a user of an online service, and is presented on the display as a result of a user of the client device viewing content provided by the online service. At completion of the period of time, the client device determines a set of remaining user indicators displayed on the display of the client device, and transmits, to a server associated with the online service, a batched request for presence information for a set of users corresponding to the set of remaining user indicators. The client device presents presence indicators indicating the status of the users based on the requested presence information.

Abstract: Contemplated systems and methods allow for prediction of time for MDS to AML transition using a predictive model that is based on selected features with significant differential expression levels and/or pathway activity between MDS to AML cells.