Reckonect SAS is a company rooted in deeptech applied to biology. We combine AI-based
information retrieval models and biomedical expertise to map disease progression from 3 perspectives: scale,
causality and context of pathological development.
We are distinct from alternatives, noteworthy because
we rely on avatars of pathologies and we can integrate heterogenous data (eg: multi-omics, semantic). See more at
reckonect.com.
Drug screening for the identification of lead and mechanism of action, is expensive
(several thousands of euros), long (3-5 years) with a high failure rate as only 4% will yield a licensed drug.
Reckonect’s AI based in silico drug screening for preclinical studies, drastically decreases both the duration
and expenses of this step.
We offer several solutions for in silico drug screening, de novo positioning,
indication repositioning or even identification of mechanism of action.
Ideally positioned upstream of the in
vitro testing you designed; you alleviate the risk of failing the in-vitro tests and save both time and money by at
least a factor of 100 folds. Â Â Â Â Â
A mechanistic disease progression model is a representation of all accessible
relevant knowledge and validated experimental, preclinical, clinical data or even real-world evidence.
It is
built upon the understanding of physiology in depth, especially on the expertise of the hierarchy of biological
events.
The most promising model of disease progression are avatars and digital twins in healthcare.
Avatars as mechanistic disease progression models, is a game changer in diagnostics
and therapeutic R&D.
They are known to drastically compress both time and expenses of basic, translational
and clinical research in diagnostics and therapeutics discovery.
They are especially relevant on the following
tasks: the identification of biomarkers and drivers of disease progression, the unravelling of the synergies or
antagonisms of comorbidities or genetic background on a given disease progression.
Mechanistic disease
progression avatars rationalize very early the R&D process with the complex physiological knowledge.
They
are currently impacting profoundly therapeutic discovery as they to in silico drug screening, de novo indication
positioning, and indication repositioning.
Disease is associated with a prototypical signature, aka several non-physiological
alterations in the behavior of biological components (e.g. genes, lipids, proteins, metabolites, signaling networks,
cell types, organs).
More precisely, this signature of biomarkers and drivers, is composed of distinct types
and identities of components (e.g. a given cellular ecosystem for which several signaling are altered) and is
specific of associated conditions such as comorbidities.
Although analysis of omics data by data scientists is still the main strategy to
perform the discovery of biomarkers and drivers, it suffers from several caveats that mechanistic models resolve.
Nowadays, mechanistic models such as avatars of disease progression, are combined with data science tools.
Mechanistic models are a kind of avatars that is now combined with data science to offer the only way of
integration of both biological cascades and physiological context, as well as the various organism levels
(molecules, cell types, organs).
This dual strategy is proven to be extremely powerful as it facilitates
empirical data integration (e.g. omics, imaging, semantics) and therefore impacts directly on candidate discovery,
qualification, verification, validation and optimization.
Disease progression encompasses several levels of organization: molecular, cellular
and tissular. When it comes to understand and target disease progression, one has to carefully consider the
existence of comorbidities or existing conditions.
In such cases, the classical statistical data science on
omics or imaging data, remains too reductionist to take into account comorbidities.
On the contrary, Reckonect
can easily build and combined mechanistic disease progression models of several conditions to identify their
potential synergies or antagonisms, using and empowering datascience workflows.
Although single cell omics analysis (transcriptomics and proteomics), is the main way
to identify cell types persity in a sample mixture, it strongly relies on the existing validated cell type gene or
protein expression signature from several omic based databases.
Unfortunately, this approach has a known
Achilles heel, since there is only very little validation of such databases for some cell types and a limited number
of omics analysis.
Reckonect built a strategy to complete missing information of signature databases and
identify cellular ecosystem composition dynamics.
An avatar of disease is a graphical overview of the pathophysiological context of the
disease.
It contains different entities and depict their spatiotemporal interactions.
We provide
different levels of description, from a systemic to a molecular one.
Each interaction is based on specific
pieces of evidence identified in the complex biomedical literature.
Let’s make it clear: like in space and energy industry, avatar usage provides you with
benefits regarding: time, precision, exhaustivity & spotting the novelty.
First, you save time
because we already compiled all sources that you would have queried manually: the associated evidences are compiled
and presented clearly and are ready to use.
Second, you gain precision because we depict the
relationships between entities at different scales, from cellular and organs long range interactions to signal
transduction. This is not only a window to the hierarchy of events that supports disease progression, this is also
acknowledging the biological context of it.
Finally, describing a disease with a given avatar is proved
to be very exhaustive. Why? because we use transversal biomedical data that more than 90% scientist would have never
accessed otherwise, stepping into knowledge gaps without acknowledging them nor identifying the sweet spot of
novelty.
On an everyday usage in a lab, avatar usage helps you to quickly make sharp choices and focus on
un-identified otherwise knowledge gaps. So, at the end of the month, you saved money and keep an eagle view to
deploy your R&D strategy.
Whatever your reasons are, our policy is simple: you do not like it, we guarantee
your refund. See? No worry.
Despite our best expertise to provide you with the best possible avatar, it
could be that you are not happy with it. Do not worry, we fully reimburse your payment. For that, you’ll just need
to fill in a survey within 24h of the payment (see survey here), to help us improve our service and understand why
you are not happy with your purchase.
There are 2 distinct versions of avatars:
Exploratory avatar.
This
avatar corresponds to a visual representation and lists of entities that occur in the disease. The Exploratory
avatar is interactive, with the sentences justifying the presence of the entities and the links between entities. In
addition, the most important entities are already ranked, regarding their relative importance for the construction
of the avatar.
You can view a Exploratory avatar as an ultra-exhaustive knowledge base that encompasses a
good chunk of the knowledge concerning a disease. In addition, it allows you to draw fair assumptions and fuel your
exploratory R&D work.
Causal avatar.
A causal avatar is an interactive map that represents the
evolution of a pathology and the relationships details. It requested in-depth evaluation of information by manual
annotation, noticeably helping you to identify candidate drivers of disease progression.
Sure, you can, and we specifically designed their structure for it.
Mixing
your data, or even publicly available data you found, with avatar, is a great and swift solution to better find and
use the information.
We have good news for you: we already offer our services to do so.
No, you cannot. Reckonect SAS solutions, are solely for R&D
usage (basic research, pre-clinical and clinical). They must not be used for patient care. If you are interested by
clinical usage, please contact us.
Reach Us
Ever since its creation, Reckonect SAS helped several non-profit organizations, pro
bono. We intend to facilitate the access to resources, training and models.
Check our dedicated open
innovation website for resources, or contact us for help :
Your account is created right after you proceeded with the payment, your access is
granted for 1 month.
Once your granted access is over, you will have the possibility to renew your
subscription (button extend by one month in tour account). To acknowledge your fidelity, we offer a 20% discount
from the original price of the avatar.
We compiled several open access biomedical resources, ranging from basic to clinical
research :
In addition, we also originally retrieved and structured disease progression facts
that were not done so before.
Once you bought your avatars, you can take advantage of the visualization and
exploration tools we provide you with. Part of the data accessible with the avatars will be downloadable on your
computer for offline consultation. Although the downloading, the reselling or providing an access to a third party
of the avatars is not granted, we can do so upon explicit written agreement and contractualization ( Reach Us
).
NLPs are fantastic tools to retrieve, structure and analyze biomedical information.
There are several kind of NLPs solutions that we used within a large spectrum of existing ones. Want to learn more
about it ?
Please check our resources center to learn more :
Exploratory avatars are not species specific, whereas Causal ones are, mainly
focusing on human data.
To search for an avatar, you have to go on the Avatars search menu, and enter the
name of the disease using your terms and synonyms. A vast array of terminology is working, including MeSH
terms.
On Exploratory avatars, you can use 2 different types of search
functions.
The first one, located above the avatars, will help you to find the entity you are looking for
in the graphical view.
The second one, located above the right panel (the list of entities) will help you
to find an entity in the list, according to the entity’s ontology.
On causal avatars, you can search for
a box name using the search box. You can also use the search bar when you are searching for a protein (in the
general list, and also when you are on a box/edge information).
First, make sure you have correctly written the entity you are looking for, by using
the appropriate term. Second, if the entity of interest still does not appear on the graph, although it is present
in the list, you can increase the number of entities present with the Node density slider. Finally, it can be that
the data, connection and term that you are thinking about, are not yet present in the avatar representation… for the
moment!
Try to vary the terms with synonyms you can think of, or even classical terminologies
as the ones provided by biomedical ontologies (eg. MeSH).
For example, if you are interested in diabetes
mellitus, type 2, you would be tempted to write T2DM, but it might not work. An alternative, however, could be
Diabetes Mellitus, Type 2, or NIDDM, or Diabetes Mellitus, Non-Insulin Dependent.
On avatars, you have a protein tab, where you can see the most important proteins.
Our proteins are ranked in order to classify them according to their importance in the disease.
Yes, you can!
There are tremendous advantages of using avatar comparison.
Case #1: you might want to identify the common features and their specificities between avatars as for instance to compare stages during pathological progression.
Case #2: you could be interested in the relationship between a given disease and its co-morbidities associations. Finding the intersection and specificities of each of the pathologies powerfully enlightens this complex relationship, by pinpointing at endocrinal cues, common signaling networks that could pave, or restrict, the road for disease progression.
Please try another web browser (e.g. chrome).
Since avatars empower R&D researchers by compressing expenses in time and
money, they are naturally used extensively on day-to-day applications. We listed 7 applications and avatar type
association. Generally, bear in mind that Exploratory Avatars are typically for exploratory research while Causal
are more for state-of-the-art and consensus like usage.
Identify new avenues and knowledge gaps to
pioneer your field
Exploratory avatars
Causal avatars, … depending on the exploratory path you might
want to take. In the first case it’ll be extremely new and fresh, yet risky, while in the second option, it’ll be
still news, but more solid yet probably more of a beaten road.
Prepare a touch point with other experts
Exploratory avatars if exploratory brainstorming and creative sessions,
Causal avatars for a fast call to
action with limited exploration.
Prepare a review for publication or for an internal note
Causal avatar
for review,
Exploratory avatar for a point of view or a hypothesis type publication.
Build preliminary
results for a grant application
Causal avatar to write down the state of the art and the introduction,
Exploratory avatar for hypothesis generation.
Rationally identify molecular or cellular signature of disease
progression
Nothing will beat a Causal avatar in that case.
Identify complementary collaborators for
trans-national calls
Exploratory avatars for exploratory and risky out of the box calls,
Causal avatars
for classical calls that are more conservative.
Shed new light and interpret fully your complex results, as
omics or semantics.
Exploratory avatars to analyze your data with candid view,
Causal avatars for a
candidate approach, and no fear of discarding out-of-the-box insights.
Access is single-user and is granted for a period of 3 months. To facilitate thinking
in common, we apply an extra fee so you can share your session with a limited number of additional users.
The duplication, the construction of similar representation or with a significative
degree of resemblance and/or composition, the downloading and re-selling of the avatar access, are all strictly
forbidden.
Avatars and associated solutions provided on Reckonect SAS website are solely
intended for R&D purpose and must not be used for patient care.
Please see our General Sales
Conditions for further details. For special inquiry that might not be adressed in GSCs, please contact us.
Please contact us.
