Abstract: This disclosure provides for a highly-efficient and scalable compression tool that compresses quality scores, preferably by capitalizing on sequence redundancy. In one embodiment, compression is achieved by smoothing a large fraction of quality score values based on k-mer neighborhood of their corresponding positions in read sequences. The approach exploits the intuition that any divergent base in a k-mer likely corresponds to either a single-nucleotide polymorphism (SNP) or sequencing error; thus, a preferred approach is to only preserve quality scores for probable variant locations and compress quality scores of concordant bases, preferably by resetting them to a default value. By viewing individual read datasets through the lens of k-mer frequencies in a corpus of reads, the approach herein ensures that compression “lossiness” does not affect accuracy in a deleterious way.

Abstract: This disclosure provides for a highly-efficient and scalable compression tool that compresses quality scores, preferably by capitalizing on sequence redundancy. In one embodiment, compression is achieved by smoothing a large fraction of quality score values based on k-mer neighborhood of their corresponding positions in read sequences. The approach exploits the intuition that any divergent base in a k-mer likely corresponds to either a single-nucleotide polymorphism (SNP) or sequencing error; thus, a preferred approach is to only preserve quality scores for probable variant locations and compress quality scores of concordant bases, preferably by resetting them to a default value. By viewing individual read datasets through the lens of k-mer frequencies in a corpus of reads, the approach herein ensures that compression “lossiness” does not affect accuracy in a deleterious way.

Abstract: A method for computation- and memory-efficient DNA sequencing. In one embodiment, the approach herein is used to facilitate genome assembly for state-of-the-art and low-error long-read data. In this embodiment, the approach herein implements a minimizer-space de Bruijn graph, which—instead of building an assembly over sequence bases (in a base-space wherein an alphabet sequence comprises nucleotide letters)—performs assembly in a minimizer-space (wherein an alphabet sequence comprises an ordered sequence of minimizers), and later converts the assembly back to base-space assemblies. Specifically, and in a preferred implementation, each read is initially converted to an ordered sequence of its minimizers. The order of the minimizers is maintained to facilitate reconstructing the entire genome as an ordered list.

Abstract: Computationally-efficient techniques facilitate secure pharmacological collaboration with respect to private drug target interaction (DTI) data. In one embodiment, a method begins by receiving, via a secret sharing protocol, observed DTI data from individual participating entities. A secure computation then is executed against the secretly-shared data to generate a pooled DTI dataset. For increased computational efficiency, at least a part of the computation is executed over dimensionality-reduced data. The resulting pooled DTI dataset is then used to train a neural network model. The model is then used to provide one or more DTI predictions that are then returned to the participating entities (or other interested parties).

Abstract: A method for efficient search of protein sequence databases for proteins that have sequence, structural, and/or functional homology with respect to information derived from a search query. The method involves transforming the protein sequences into vector representations and searching in a vector space. Given a database of protein sequences and a learned embedding model, the embedding model is applied to each amino acid sequence to transform it into a sequence of vector representations. A query sequence is also transformed into a sequence of vector representations, preferably using the same learned embedding model. Once the query has been embedded in this manner, proteins are retrieved from the database based on distance between the query embedding and the protein embeddings contained within the database. Rapid and accurate search of the vector space is carried out using exact search using metric data structures, or approximate search using locality sensitive hashing.

Abstract: A method of compressive read mapping. A high-resolution homology table is created for the reference genomic sequence, preferably by mapping the reference to itself. Once the homology table is created, the reads are compressed to eliminate full or partial redundancies across reads in the dataset. Preferably, compression is achieved through self-mapping of the read dataset. Next, a coarse mapping from the compressed read data to the reference is performed. Each read link generated represents a cluster of substrings from one or more reads in the dataset and stores their differences from a locus in the reference. Preferably, read links are further expanded to obtain final mapping results through traversal of the homology table, and final mapping results are reported. As compared to prior techniques, substantial speed-up gains are achieved through the compressive read mapping technique due to efficient utilization of redundancy within read sequences as well as the reference.

Abstract: A method for efficient search of protein sequence databases for proteins that have sequence, structural, and/or functional homology with respect to information derived from a search query. The method involves transforming the protein sequences into vector representations and searching in a vector space. Given a database of protein sequences and a learned embedding model, the embedding model is applied to each amino acid sequence to transform it into a sequence of vector representations. A query sequence is also transformed into a sequence of vector representations, preferably using the same learned embedding model. Once the query has been embedded in this manner, proteins are retrieved from the database based on distance between the query embedding and the protein embeddings contained within the database. Rapid and accurate search of the vector space is carried out using exact search using metric data structures, or approximate search using locality sensitive hashing.

Abstract: An activity of interest is modeled by a non-stationary discrete stochastic process, such as a pattern of mutations across a cancer genome. Initially, input genomic data is used to train a model to predict rate parameters and their associated uncertainty estimation for each of a set of process regions. For any arbitrary set of indexed positions of the stochastic process that are identified in an information query, the rate parameters and their associated estimation uncertainties are scaled using the model to obtain a distribution of the events of interest and their associated estimation uncertainties for the set of indexed positions. In one practical application, and in response to a search query associated with one or more base-pairs, a result is then returned. The result, which represents deviations between the estimated and observed mutation rates, is used to identify genomic elements that have more mutations than expected and therefore constitute previously unknown driver mutations.

Abstract: Computationally-efficient techniques facilitate secure pharmacological collaboration with respect to private drug target interaction (DTI) data. In one embodiment, a method begins by receiving, via a secret sharing protocol, observed DTI data from individual participating entities. A secure computation then is executed against the secretly-shared data to generate a pooled DTI dataset. For increased computational efficiency, at least a part of the computation is executed over dimensionality-reduced data. The resulting pooled DTI dataset is then used to train a neural network model. The model is then used to provide one or more DTI predictions that are then returned to the participating entities (or other interested parties).

Abstract: A method for efficient search of protein sequence databases for proteins that have sequence, structural, and/or functional homology with respect to information derived from a search query. The method involves transforming the protein sequences into vector representations and searching in a vector space. Given a database of protein sequences and a learned embedding model, the embedding model is applied to each amino acid sequence to transform it into a sequence of vector representations. A query sequence is also transformed into a sequence of vector representations, preferably using the same learned embedding model. Once the query has been embedded in this manner, proteins are retrieved from the database based on distance between the query embedding and the protein embeddings contained within the database. Rapid and accurate search of the vector space is carried out using exact search using metric data structures, or approximate search using locality sensitive hashing.

Abstract: An activity of interest is modeled by a non-stationary discrete stochastic process, such as a pattern of mutations across a cancer genome. Initially, input genomic data is used to train a model to predict rate parameters and their associated uncertainty estimation for each of a set of process regions. For any arbitrary set of indexed positions of the stochastic process that are identified in an information query, the rate parameters and their associated estimation uncertainties are scaled using the model to obtain a distribution of the events of interest and their associated estimation uncertainties for the set of indexed positions. In one practical application, and in response to a search query associated with one or more base-pairs, a result is then returned. The result, which represents deviations between the estimated and observed mutation rates, is used to identify genomic elements that have more mutations than expected and therefore constitute previously unknown driver mutations.

Abstract: An activity of interest is modeled by a non-stationary discrete stochastic process, such as a pattern of mutations across a cancer genome. Initially, input genomic data is used to train a model to predict rate parameters and their associated uncertainty estimation for each of a set of process regions. For any arbitrary set of indexed positions of the stochastic process that are identified in an information query, the rate parameters and their associated estimation uncertainties are scaled using the model to obtain a distribution of the events of interest and their associated estimation uncertainties for the set of indexed positions. In one practical application, and in response to a search query associated with one or more base-pairs, a result is then returned. The result, which represents deviations between the estimated and observed mutation rates, is used to identify genomic elements that have more mutations than expected and therefore constitute previously unknown driver mutations.

Abstract: A method of viral escape profiling is used in association with antiviral or vaccine development. The method begins by training a language-based model against training data comprising a corpus of viral protein sequences of a given viral protein to model a viral escape profile. The viral escape profile represents, for one or more regions of the given viral protein, a relative viral escape potential of a mutation, the relative viral escape potential being derived as a function that combines both “semantic change,” representing a degree to which the mutation is recognized by the human immune system (i.e., antigenic change), and “grammaticality,” representing a degree to which the mutation affects viral infectivity (i.e. viral fitness). Using the model, a region of the given viral protein having an escape potential of interest is identified. Information regarding the region is then output to a vaccine or anti-viral therapeutic design and development workflow.

Abstract: Computationally-efficient techniques facilitate secure crowdsourcing of genomic and phenotypic data, e.g., for large-scale association studies. In one embodiment, a method begins by receiving, via a secret sharing protocol, genomic and phenotypic data of individual study participants. Another data set, comprising results of pre-computation over random number data, e.g., mutually independent and uniformly-distributed random numbers and results of calculations over those random numbers, is also received via secret sharing. A secure computation then is executed against the secretly-shared genomic and phenotypic data, using the secretly-shared results of the pre-computation over random number data, to generate a set of genome-wide association study (GWAS) statistics. For increased computational efficiency, at least a part of the computation is executed over dimensionality-reduced genomic data.

Abstract: A method of compressive read mapping. A high-resolution homology table is created for the reference genomic sequence, preferably by mapping the reference to itself. Once the homology table is created, the reads are compressed to eliminate full or partial redundancies across reads in the dataset. Preferably, compression is achieved through self-mapping of the read dataset. Next, a coarse mapping from the compressed read data to the reference is performed. Each read link generated represents a cluster of substrings from one or more reads in the dataset and stores their differences from a locus in the reference. Preferably, read links are further expanded to obtain final mapping results through traversal of the homology table, and final mapping results are reported. As compared to prior techniques, substantial speed-up gains are achieved through the compressive read mapping technique due to efficient utilization of redundancy within read sequences as well as the reference.

Abstract: A method of compressive read mapping. A high-resolution homology table is created for the reference genomic sequence, preferably by mapping the reference to itself. Once the homology table is created, the reads are compressed to eliminate full or partial redundancies across reads in the dataset. Preferably, compression is achieved through self-mapping of the read dataset. Next, a coarse mapping from the compressed read data to the reference is performed. Each read link generated represents a cluster of substrings from one or more reads in the dataset and stores their differences from a locus in the reference. Preferably, read links are further expanded to obtain final mapping results through traversal of the homology table, and final mapping results are reported. As compared to prior techniques, substantial speed-up gains are achieved through the compressive read mapping technique due to efficient utilization of redundancy within read sequences as well as the reference.

Abstract: A method of viral escape profiling is used in association with antiviral or vaccine development. The method begins by training a language-based model against training data comprising a corpus of viral protein sequences of a given viral protein to model a viral escape profile. The viral escape profile represents, for one or more regions of the given viral protein, a relative viral escape potential of a mutation, the relative viral escape potential being derived as a function that combines both “semantic change,” representing a degree to which the mutation is recognized by the human immune system (i.e., antigenic change), and “grammaticality,” representing a degree to which the mutation affects viral infectivity (i.e. viral fitness). Using the model, a region of the given viral protein having an escape potential of interest is identified. Information regarding the region is then output to a vaccine or anti-viral therapeutic design and development workflow.

Abstract: Computationally-efficient techniques facilitate secure crowdsourcing of genomic and phenotypic data, e.g., for large-scale association studies. In one embodiment, a method begins by receiving, via a secret sharing protocol, genomic and phenotypic data of individual study participants. Another data set, comprising results of pre-computation over random number data, e.g., mutually independent and uniformly-distributed random numbers and results of calculations over those random numbers, is also received via secret sharing. A secure computation then is executed against the secretly-shared genomic and phenotypic data, using the secretly-shared results of the pre-computation over random number data, to generate a set of genome-wide association study (GWAS) statistics. For increased computational efficiency, at least a part of the computation is executed over dimensionality-reduced genomic data.

Abstract: The redundancy in genomic sequence data is exploited by compressing sequence data in such a way as to allow direct computation on the compressed data using methods that are referred to herein as “compressive” algorithms. This approach reduces the task of computing on many similar genomes to only slightly more than that of operating on just one. In this approach, the redundancy among genomes is translated into computational acceleration by storing genomes in a compressed format that respects the structure of similarities and differences important to analysis. Specifically, these differences are the nucleotide substitutions, insertions, deletions, and rearrangements introduced by evolution. Once such a compressed library has been created, analysis is performed on it in time proportional to its compressed size, rather than having to reconstruct the full data set every time one wishes to query it.

Abstract: The redundancy in genomic sequence data is exploited by compressing sequence data in such a way as to allow direct computation on the compressed data using methods that are referred to herein as “compressive” algorithms. This approach reduces the task of computing on many similar genomes to only slightly more than that of operating on just one. In this approach, the redundancy among genomes is translated into computational acceleration by storing genomes in a compressed format that respects the structure of similarities and differences important to analysis. Specifically, these differences are the nucleotide substitutions, insertions, deletions, and rearrangements introduced by evolution. Once such a compressed library has been created, analysis is performed on it in time proportional to its compressed size, rather than having to reconstruct the full data set every time one wishes to query it.