GeneFlow 2.0 is a connected workspace for mitochondrial DNA analysis. It is designed to reduce the handoffs between file storage, sequence quality review, classification, phylogenetics, and reporting.
The difficult part of a mitochondrial DNA project is rarely one command. It is the chain around that command: receiving files from an instrument or collaborator, matching them to sample identifiers, deciding whether sequence quality is acceptable, running classification consistently, explaining an unexpected result, rebuilding a tree, and preparing material for review.
That chain often lives in several places at once. Raw files sit on a shared drive. Quality observations are copied into a spreadsheet. An analyst runs a tool from a terminal and saves output under a new filename. Parameters are described later in a notebook. A PDF is emailed to a collaborator. Each handoff creates another opportunity to lose context or compare results produced from different inputs.
GeneFlow 2.0 brings those activities into a project-centered workflow for mtDNA research. The goal is not to remove scientific judgment. It is to give that judgment a clearer evidence trail.
Start with the source, not the spreadsheet
The Data Vault accepts AB1, FASTA, FASTQ, and VCF files. On upload, GeneFlow records the original name, file type, byte size, uploader, storage key, processing status, and a SHA-256 checksum calculated from the uploaded bytes. Files are stored outside the relational database, while their operational metadata remains attached to the project.
Parsing turns supported inputs into project samples. An AB1 file contributes base calls, Phred scores, trace channels, instrument metadata when available, and a link back to its source file. FASTA records become sequence-bearing samples. Consensus-scale FASTQ records retain sequence and quality values. VCF sample columns become sample records linked to the uploaded variant file.
This matters because a sample is no longer only a row named sample-17. It can remain connected to the file from which GeneFlow derived it and to the analyses that follow.
Review quality where the sequence lives
For AB1 traces and supported FASTQ records with Phred scores, GeneFlow computes quality summaries, identifies low-quality regions, assigns an overall quality category, and proposes left and right trim coordinates. For Sanger data, researchers can inspect the chromatogram evidence rather than accepting a single pass/fail label without context.
A trim suggestion is still a suggestion. Primer sequence, mixed peaks, contamination, heteroplasmy, strand coverage, and the study’s acceptance criteria require scientific review. GeneFlow makes the quantitative evidence easier to inspect; it does not turn a quality threshold into a universal biological decision.
Move from sequence to classification
GeneFlow’s current automated haplogroup workflow uses HaploGrep3. Sequence-bearing samples are classified from their consensus or base calls, while VCF-derived samples can be classified from the source VCF. The resulting record can include the called haplogroup, confidence value, found mutations, status, and raw tool output. A sample-level consensus record keeps the review state and can retain resolution notes when human interpretation is needed.
Tool availability and suitable input still matter. A failed classifier, missing sequence, malformed input, or low-confidence result remains visible as a status to investigate rather than becoming a silently missing row.
Define and observe repeatable runs
GeneFlow pipelines describe QC and haplogroup steps as ordered or branching graphs. A definition records its project, graph, creator, and selected rCRS or RSRS reference setting. Each execution receives its own run status, timing, creator, input overrides, and error field. Per-step records retain step type, inputs, outputs, timing, status, and errors.
That structure helps answer practical questions: Which run failed? Which sample identifier was supplied to a step? Did an independent branch complete? Was this output produced before or after a pipeline definition changed?
Pipeline records do not capture every property of the surrounding compute environment. Laboratories that require exact environment reconstruction should also govern tool versions, reference assets, container or package versions, and deployment configuration in their SOPs.
Build project-level context
For projects with at least three sequence-bearing samples, GeneFlow can align sequences with MAFFT and infer a Newick tree with FastTree. Tree records retain status, method, tip labels, sequence count, creator, and an error when inference cannot complete.
Reports bring selected QC, haplogroup, phylogenetic, or full-project information into structured records and generated files. Reports belong to a project, can be sample-specific, and can optionally be related to a pipeline run. That explicit relationship is more useful than treating every exported PDF as self-explanatory, but teams should still record interpretation, approval, and publication decisions according to their own governance process.
Collaborate without sharing one account
Projects are also the access boundary. Every project membership has one of four roles: owner, admin, member, or viewer. All active members can read project material. Owners, admins, and members can create or update it; deletion is limited to owners and admins. GeneFlow also prevents removal of a project’s last owner.
Those roles support routine separation between contributors and reviewers. They do not replace institutional identity management, access reviews, backup policy, or rules governing which data may be placed in a project.
What GeneFlow 2.0 is, and is not
GeneFlow 2.0 is focused on current mtDNA workflows. It is not a general nuclear-genome analysis environment, a raw-read assembler, a notebook platform, or a clinical interpretation system. FASTQ support is intended for consensus-scale data; raw read sets above the platform’s supported record limit require assembly elsewhere before upload. Analysis outputs must be reviewed by qualified researchers in the context of the study.
The platform records history for selected entities, including projects, samples, pipeline definitions, and haplogroup consensus records. That is useful provenance, but it is not a claim of immutable storage or a complete security audit log. Production deployments can use S3-compatible object storage, but backup, versioning, retention, encryption, logging, and recovery remain controls that each operator must configure and test.
This release is about making the path from source file to reviewable result easier to follow. Files, samples, QC, classification, pipeline execution, trees, reports, and project membership now share a common working context. Researchers still make the decisions. GeneFlow helps keep the evidence for those decisions connected.
The GeneFlow Team