Systems and methods for analyzing viral nucleic acids
The invention provides systems and methods for analyzing viruses by representing viral genetic diversity with a directed acyclic graph (DAG), which allows genetic sequencing technology to detect rare variations and represent otherwise difficult-to-document diversity within a sample. Additionally, a host-specific sequence DAG can be used to effectively segregate viral nucleic acid sequence reads from host sequence reads when a sample from a host is subject to sequencing. Known viral genomes can be represented using a viral reference DAG and the viral sequence reads from the sample can be compared to viral DAG to identify viral species or strains from which the reads were derived. Where the viral sequence reads indicate great genetic diversity in the virus that was infecting the host, those reads can be assembled into a DAG that itself properly represents that diversity.
1. A method for analyzing a virus, the method comprising using at least one hardware processor connected to a tangible memory subsystem to perform:
obtaining a plurality of sequence reads from a sample containing viral nucleic acid;
creating, in the tangible memory subsystem, a viral DNA reference graph representing a plurality of known viral sequences, the viral DNA reference graph comprising a directed acyclic graph (DAG) comprising a plurality of vertices stored as objects in the tangible memory subsystem, wherein sequence strings of the known viral sequences that match each other when aligned are each represented by a single object and sequence strings that vary are represented as alternate objects, wherein at least one sequence string comprises a plurality of symbols, and wherein each object is stored in the tangible memory subsystem as a sequence string and a list of one or more pointers to adjacent objects, wherein each pointer identifies a physical location in the tangible memory subsystem at which an adjacent object is stored, such that the objects are linked to represent each of the known viral sequences as a path through the viral DNA reference graph;
aligning the sequence reads to paths through the viral DNA reference graph, wherein the aligning comprises calculating match scores between a sequence read and a sequence string associated with each vertex, and looking backwards at each vertex to predecessor vertices if and only if a symbol comprises the first symbol of the sequence string associated with its vertex to select a path with an optimal score;
retrieving, based on the aligned sequence reads to paths through the viral DNA reference graph, an identity of a virus for the one of the plurality of known viral sequences; and
providing a report that includes the identity of the virus.
2. The method of claim 1 , wherein creating a viral DNA reference graph further comprises:
obtaining the plurality of known viral sequences from an online sequence database;
finding and deleting redundancies among homologous portions of the known viral sequences, leaving the segments of the known viral sequences that match each other when aligned and segments of the known viral sequences that vary; and
creating one of the objects in the tangible memory subsystem for each of the segments.
3. The method of claim 1 , wherein the directed graph represents a plurality of known variations in a viral genome, in which each of the known variations is associated with a path through the directed graph.
4. The method of claim 1 , further comprising retrieving, based on the alignments, an identity of a plurality of viruses for the sample.
5. The method of claim 4 , wherein the report includes a list of viral species or strains determined based on the retrieved identities, and the report further quantifies an amount of each viral species or strain present in the viral nucleic acid.
6. The method of claim 4 , wherein the known viral sequences are of the same species as the viral nucleic acid, the method further comprising characterizing a quasispecies of the virus in the sample.
7. The method of claim 6 , wherein the report includes a list of strains of the virus and a graphical representation of a graph of aligned sequence reads from the sample.
8. The method of claim 1 , further comprising:
obtaining sequence reads from genomic nucleic acid from a healthy cell of a subject;
aligning the sequence reads to a human reference graph;
identifying, based on the aligned sequence reads, a plurality of variants;
creating a subject-specific reference graph that includes the plurality of variants;
obtaining disease sequence reads from a diseased cell of the subject and aligning them to the subject-specific reference graph, wherein disease sequence reads that do not align and disease sequence reads that contain variants relative to the subject-specific reference DAG are considered candidate reads; and
aligning the candidate reads to the directed graph representing known viral sequences.
9. The method of claim 8 , wherein the report includes the identities of one or more viruses identified by aligning the candidate reads to the directed graph representing known viral sequences.
10. The method of claim 1 , wherein the sequence from the sample comprises a sequence read created by sequencing the viral nucleic acid using a sequencing instrument.
11. The method of claim 1 , further comprising depleting host-specific nucleic acids from the sample.
12. The method of claim 1 , wherein objects of the viral DNA reference graph comprise vertex objects connected by edge objects and an adjacency list for each vertex object and edge object, wherein the adjacency list for a vertex object or edge object lists the edge objects or vertex objects to which that vertex object or edge object is adjacent, wherein each entry in an adjacency list is a pointer to the adjacent vertex object or edge object, wherein each pointer identifies a physical location in the memory subsystem at which the adjacent object is stored.
13. The method of claim 1 , wherein the viral DNA reference graph uses index-free adjacency to link the objects into paths to represent the plurality of known viral sequences.
14. The method of claim 1 , wherein converting the sequence into the alignment comprises using the processor to perform a multi-dimensional look-back operation to find a highest-scoring trace through a multi-dimensional matrix.
15. The method of claim 1 , wherein creating the viral DNA reference graph comprises:
associating an initial object with a sequence string representing one of the known viral sequences;
dividing the initial object into multiple objects at positions in which known variations from the plurality of known viral sequences occur; and
connecting the plurality of objects to create paths such that there is a path for each known variation.
16. A system for analyzing a virus, the system comprising a processor coupled to a tangible memory subsystem, wherein the tangible memory subsystem contains instructions that when executed by the processor cause the system to perform:
obtaining a plurality of sequence reads from a sample containing viral nucleic acid;
creating, in the tangible memory subsystem, a viral DNA reference graph representing a plurality of known viral sequences, the viral DNA reference graph comprising a directed acyclic graph (DAG) comprising a plurality of vertices stored as objects in the tangible memory subsystem, wherein sequence strings of the known viral sequences that match each other when aligned are each represented by a single object and sequence strings that vary are represented as alternate objects, and wherein each object is stored in the tangible memory subsystem as a sequence string and a list of one or more pointers to adjacent objects, wherein each pointer identifies a physical location in the tangible memory subsystem at which an adjacent object is stored, such that the objects are linked to represent each of the known viral sequences as a path through the viral DNA reference graph;
aligning the sequence reads to paths through the viral DNA reference graph, wherein the aligning comprises calculating match scores between a sequence read and a sequence string associated with each vertex, and looking backwards at each vertex to predecessor vertices if and only if a symbol comprises the first symbol of the sequence string associated with its vertex to select a path with an optimal score;
retrieving, based on the aligned sequence reads to paths through the viral DNA reference graph, an identity of a virus for each of the plurality of sequences; and
providing, via an input/output device coupled to the processor, a report that includes the retrieved identities.
17. The system of claim 16 , further operable to:
obtain the plurality of known viral sequences from an online sequence database;
find and delete redundancies among homologous portions of the known viral sequences, leaving the segments of the known viral sequences that match each other when aligned and segments of the known viral sequences that vary; and
create one of the objects in the tangible memory subsystem for each of the segments.
18. The system of claim 16 , wherein host-derived sequence reads are removed from the sample prior to finding alignments.
19. The system of claim 16 , further operable to:
obtain sequence reads from genomic nucleic acid from a healthy cell of a subject;
align the sequence reads to a human reference graph;
identify, based on the aligned sequence reads, a plurality of variants;
create a subject-specific reference graph that includes the plurality of variants;
obtain tumor-associated sequence reads from a tumor of the subject;
align the tumor-associated sequence reads to the subject-specific reference graph; and
align tumor-associated reads that do not align to the subject-specific reference graph to the directed graph representing a plurality of known viral sequences.
20. The system of claim 16 , wherein the report includes a list of viral species or strains determined based on the retrieved identities, and the report further quantifies an amount of each viral species or strain present in the viral nucleic acid.