ishanu chattopadhyay

Asst Professsor of Medicine

ishanu@uchicago.edu

ishanu@paraknowlede.ai

Emergenet

Stamping out pandemics before the first human infection

Can we predict future mutations?

Can we define the "edge of emergence"?

Can we design countermeasures that are future proof?

Digital Twins for complex systems

Chattopadhyay, Ishanu, Kevin Wu, Jin Li, and Aaron Esser-Kahn. "Emergenet: Fast Scalable Pandemic Risk Assessment of Influenza A Strains Circulating In Non-human Hosts." (2023). Under Review in Nature

PREEMPT

curl -X POST https://us-central1-pkcsaas-01.cloudfunctions.net/emergenet_predict?api_key=7eea9f70d79c408f2b69847d6911303c -H "Content-Type: application/json" -d '{
"HA":"MKAILLVLLHTFAATSADTICVGYHANNSTDTVDTVLEKNVTVTHSVNLLEDKHNGKLCKLRGKAPLYLGKCNIAGWLLGNPECELPLTVSSWSYIVETSDSDNGTCYPGDFTNYEELREQLSSVSSFERFEMFPKESSWPNHETNKSVTAACPYAGASSFYRNLIWLVKKDDSYPMLNISYVNNKGKEVLVLWGIHHPPTEDDQKWLYKNADAYVFVGTSTYSQKFEPEIATRPRVRDQTGRMNYYWTLVKPGDKITFEATGNLVVPRYAFAMNRGSESGIIISDAPVHDCNTICQTPKGALNTSLPFQNVHPVTIGECPKYIKSTRLKMATGLRNTPSIQSRGLFGAIAGFIEGGWTGMVDGWYGYHHQNEQGSGYAADQKSTQRAVDGITNKVNSIIERMNSQFTAVGKEFSNLERRIENLNKKVDDGFLDVWTYNAELLILLENERTLDFHDSNVKNLYERVRNQLRNNAKEIGNGCFEFYHKCDNTCMESVKNGTYDYPKYSEESKLNREEIDGVKLDSTKVYQILAIYSTVASSLVVLVSLGALSFWMCSNGSL",
"NA":"MNPNQKIITIGSVSLIIATICFLMQIAILVTTVTLHFKQHDCNSSSNNQVMLCEPIIIERNKTEIVYLTNTTVEKEICPKPAEYRNWSKPQCNIAGFAPFSKDNSIRLSAGGDIWVTREPYVSCDLDKCYQFALGQGTTLNNRHSNDTVHDRTPYRTLLMNELGVPFHLGTRQVCVAWSSSSCHDGKAWLHVCITGDDNNATASFIYNGRLVDSVVSWSKNILRTQESECVCINGTCTVVMTDGSASGKADTRILFIVEGKIIHVSKLSGSAQHVEECSCYPRYPGVRCVCRDNWKGSNRPIVDINMKDYSIVSSYVCSGLVGDTPRKTDSLSSSNCLDPNNEEGDHGVKGWAFDDGDDVWMGRTINETLRLGYETFKVIKGWSKPNSKLQTNRQVIVKGGNRSGYSGIFSVEGKNCINRCFYVELIRGRREETRVWWTSNSIVVFCGTSGTYGTGSGPDGADINLMPI"
}'

cloud function emerge_predict

{"IRAT_emergence":6.599316640549309,"IRAT_impact":6.606007439884266}

\Phi_i:\prod_{j \neq i} \Sigma_j \rightarrow \mathcal{D}(\Sigma_i)

E-Net

recursive forest

E-distance

a biologically informed, adaptive distance between strains

\theta(x,y) \triangleq \\ \mathbf{E}_i \left ( \mathbb{J}^{\frac{1}{2}} \left (\Phi_i^P(x_{-i}) , \Phi_i^Q(y_{-i})\right ) \right )

This distance is "special"

smaller distances imply a quatitatively high probability of spontaneous jump

$$J \textrm{ is the Jensen-Shannon divergence }$$

Sanov's Theorem & Pinsker's Inequality

Theorem

\left \vert \ln \frac{Pr(x \rightarrow y ) }{Pr( y \rightarrow y)} \right \vert \leqq \beta \theta(x,y)

\left \vert \ln \frac{Pr(x_a \rightarrow x_h ) }{Pr( x_h \rightarrow x_h)} \right \vert \approx 0 \\ \Rightarrow Pr(x_a \rightarrow x_h ) \approx Pr(x_h \rightarrow x_h ) \\ \color{green}\Rightarrow Pr(x_a \rightarrow x_h ) \approx 1

human strain $x_{h}$ is "well-adapted" $\Rightarrow Pr(x_h\rightarrow x_h) \approx 1 $

For animal strain $x_{a}$, $ \displaystyle \theta(x_{a},x_{h}) \approx 0 $

Sanov's Theorem & Pinsker's Inequality