Datasets by developmental stage
Every VFB dataset grouped by the developmental stage of the animal it was collected from.
VFB’s datasets are documented by technique and by source elsewhere on this site, but nothing groups them by the stage of the animal they came from. That gap has come up in practice: it is why a question like “how many larval datasets does VFB hold, and which is newest?” has no direct answer on the site today, and can only be approximated by searching dataset names for the word “larval” — a search that misses anything not named that way, and catches nothing at all for embryo or pupal material.
This page is that missing reference. It lists all 210 dataset entities currently in the
VFB knowledge base (DataSet individuals — the same list search_terms returns when
filtered to dataset), grouped by developmental stage, with a link to each dataset’s own
page where a stage-agnostic reader would want to jump straight to the data.
How this was built. VFB does not currently record a structured “stage” field on
dataset entities — this table was compiled by hand from each dataset’s name, description
and publication, cross-checked against a more reliable structured signal where one
exists: which template a dataset’s images are registered to.
Registration is stage-specific — the L1 CNS template (Seymour, VFB_00050000) only ever
receives first-instar larval images, the L3 CNS template (Wood2018, VFB_00049000) only
third-instar images — so a dataset’s AlignedDatasets membership on those templates is
firmer evidence than its name. That check changed three classifications from the first,
label-only pass: the TrumanWood2018/TrumanWood2018public Truman Larval Flip-Out
Collection and SplitMeissner2024 are all registered to the L3 template rather than an
unspecified larval stage or, in Meissner’s case, no stage at all. This is a documentation
convenience, not a data-model change — nothing here is queryable from the API. A dataset
missing from a future data release, or one whose stage was misjudged, should be corrected
here directly; if VFB’s data model gains a real structured stage field later, this page
should be regenerated from it instead of maintained by hand.
Embryo
Approximately 0–22 hours after egg laying, from fertilisation to hatching. The two datasets here are single-cell/single-nucleus transcriptomic surveys of the whole embryo; VFB holds no embryonic imaging data at this time.
| Dataset | Short form |
|---|---|
| Single-cell RNA-seq study of gastrulating embryos | FBlc0006191 |
| The continuum of Drosophila embryonic development at single-cell resolution | FBlc0007797 |
Larva
The larval period runs from hatching to pupariation and is divided into three instars — L1, L2 and L3 — separated by moults. VFB’s larval holdings cluster almost entirely at the first and third instars; no dataset here is recorded as L2. Anatomy differs enough between instars that a “larval” dataset with no instar stated should not be assumed comparable to one that states it, particularly for connectomic data.
L1 (first instar)
The stage most larval EM connectomics is drawn from: the larval CNS is small enough for
a first-instar animal to be reconstructed at synaptic resolution in a single study. Most
entries below are single-paper EM reconstructions built on the l1em connectome
(Ohyama et al. 2015 and the studies that extended it) and registered to the L1 CNS
template.
L3 (third instar)
The last and largest larval instar, imaged shortly before pupariation. VFB’s L3 material
is registered to the Wood2018 L3 CNS template and includes the Truman Larval Flip-Out
Collection (TrumanWood2018, published 2018) — currently the newest of VFB’s larval
datasets by publication year.
| Dataset | Short form |
|---|---|
| L3 Larval CNS Template (Truman2016) | Truman2016 |
| L3 neuropils (WoodHartenstein2018) | WoodHartenstein2018 |
| Split-GAL4 lines from Meissner et al., 2024 | SplitMeissner2024 |
| Truman Larval Flip-Out Collection | TrumanWood2018 |
| Truman Larval Flip-Out Collection | TrumanWood2018public |
Larva (instar not stated)
Larval by name, publication or genetic targeting, but without enough evidence in VFB to assign a specific instar — none of these are registered to either larval template, most likely because they are driver-line metadata without their own aligned images. Treat the instar as unknown rather than assuming L1 or L3.
| Dataset | Short form |
|---|---|
| MCFO images of GMR-GAL4 lines from Jovanic et al., 2019 | Gen1MCFOJovanic2019 |
| Single-cell RNA-seq study of larval optic lobes | FBlc0006404 |
| Split-GAL4 lines from Jovanic et al., 2019 | SplitJovanic2019 |
| Split-GAL4 lines from Takagi et al., 2017 | SplitTakagi2017 |
Pupa
Metamorphosis, from pupariation to eclosion. Both datasets here are transcriptomic surveys of the optic lobe spanning the pupal-to-adult transition; VFB holds no pupal-only imaging data.
| Dataset | Short form |
|---|---|
| Single-cell RNA-seq study of pupal and adult optic lobes | FBlc0005659 |
| Single-cell RNA-seq study of the pupal optic lobe | FBlc0006237 |
Adult
By far VFB’s largest holding: 170 of the 210 dataset entities. Almost everything here is light-microscopy driver-line and split-GAL4 material from FlyLight, VDRC, FlyCircuit and individual labs, plus the per-paper EM connectome datasets (FAFB, hemibrain and their derivatives) and the adult-focused Fly Cell Atlas / Aging Fly Cell Atlas scRNAseq series. Sex is usually stated for the EM connectomes (FAFB and hemibrain are each a single traced female brain; MANC and Male-CNS are male) and usually unstated for driver-line collections, which typically pool both sexes.
Brain
The overwhelming majority of VFB’s adult data: everything above that targets or images central brain and optic lobe circuitry, without a more specific VNC, leg or whole-body grouping below.
VNC (ventral nerve cord)
Datasets specific to the adult ventral nerve cord — MANC and FANC on the EM side, the Court VNC/VNS templates and neuropils, and the handful of driver-line and scRNAseq sets built around VNC circuits (walking, courtship song, neck motor control).
| Dataset | Short form |
|---|---|
| Adult VNC neuropils (Court2020) | Court2020 |
| Adult VNS neuropils (Court2017) | Court2017 |
| Full VNC EM connectome, dense reconstruction (adult male) | Takemura2023 |
| Full VNS EM, sparse reconstruction (adult female) | Maniates_Selvin2020 |
| Single-cell RNA-seq study of adult ventral nerve cords | FBlc0005603 |
| Split-GAL4 lines from Bidaye et al., 2014 | SplitBidaye2014 |
| Split-GAL4 lines from Bidaye et al., 2020 | SplitBidaye2020 |
| Split-GAL4 lines from Gorko et al., 2024 | SplitGorko2024 |
| Split-GAL4 lines from Lillvis et al., 2024 | SplitLillvis2024 |
Leg, head and thorax
Peripheral and body-wall datasets that are neither brain nor VNC proper: leg proprioception and imaging, proboscis motor neurons, and thoracic scRNAseq.
| Dataset | Short form |
|---|---|
| Biomechanical origins of proprioceptive maps in the Drosophila leg | Mamiya2022 |
| GAL4 lines from McKellar et al., 2020 | McKellar2020 |
| Images of proboscis muscles from McKellar et al., 2020 | HeadMusclesMcKellar2020 |
| Millimeter-scale imaging of a Drosophila leg at single-neuron resolution | Kuan2020 |
| Single-cell RNA-seq study of the adult thorax | FBlc0006361 |
| Split-GAL4 lines from Tuthill et al., 2013 | SplitTuthill2013 |
Full CNS and whole body
Resources that deliberately span more than one region: whole-CNS connectomes (BANC, Male-CNS), descending-neuron driver-line sets that by design run from the brain into the VNC, and the Fly Cell Atlas / Aging Fly Cell Atlas scRNAseq series, which sample the whole animal rather than the nervous system alone.
Stage-agnostic resources
A small number of dataset entities in VFB’s knowledge base are reference resources rather than material from an animal at a particular stage — reused across stages by design. None fall in this bucket at the time of writing (VFB’s templates and neuropil domain sets are all stage-specific and appear above under their own stage), but the category is kept here for anything added later that should not be forced into a stage it does not have.