Lantern Pharma Inc. · NASDAQ:LTRN
Lantern Pharma — Investor Presentation (January 2026)
34 slides from Lantern Pharma's investor presentation, Jan 23, 2026
Slide 1: Panna Sharma, CEO & President, Lantern Pharma
Panna Sharma | —— . CEO & President ‘a Lantern Pharma Inc. .
Slide 2: Precision Medicine Platform
Precision Medicine Platform
Slide 3: Slide 3
Slide 4: Solving the most challenging problems using our AI for good
; . J | Wf ‘ th — Solving the most challenging problems...Using our Alfor good. ~—
Slide 5: Rare cancers...complex, challenging, economically impossible, highly fractured
Rare cancers...complex, challenging, economically impossible, highly fractured. 1 out of 5 cancer diagnosis is a rare cancer and almost 30% of all annual cancer deaths.
Slide 6: Can we do this at hyper-scale?
Can we do this at hyper-scale? Faster Deeper Y Industrialized co Democratized Qe % @ : Leveraging Al & Always Fraction of Precision Oncology 0 Available Cost to expedite therapies for rare Accelerated Clinical and underserved indications. Development & ~ Patient Access Highly Iterative Lantern Pharm
Slide 7: Rare Cancers Are Rising Globally
Rare Cancers Are Rising Globally A growing global health challenge with increasing incidence across diverse populations. € a = SO Y Increased Detection & Awareness
Slide 8: Increased Detection & Awareness
Increased Detection & Awareness ...but therapies are not advancing fast enough
Slide 9: withZETA
@ withZETA - WORLD'S TOP MINDS — “ie cuce BIOLOGISTS MANUFACTURING PREDICTIONS DOSING & SAFETY *. a REASONING CONSTANTLY LEARNING & MAKING CONNECTIONS..: IMPROVING CURES & CREATING NEW THERAPIES. This is a roadmap for the future of drug development. GENOMICS
Slide 10: Kishor, Chief Scientific Officer
Kishor Chief Scientific Officer Lantern Pharma Formerly of: National Cancer Institute, Bethesda Children’s Cancer Center, Riyadh Tata Memorial Cancer Center, Mumbai
Slide 11: A never tested compound (LP-184) was linked to an ultra-rare cancer (ATRT)
A never tested compound (LP-184) was linked to an ultra-rare cancer (ATRT) by reasoning across 4 different databases with sparse datasets and was predicted to be highly brain penetrant. IS THIS REAL ORA HALLUCINATION? ae %
Slide 12: A Devastating Pediatric Cancer With Limited Therapeutic Options
—A Devastating Pediatric Cancer With Limited Therapeutic Options ¢ FDA Orphan Drug Designation ¢ FDA Rare Pediatric Disease Designation ¢ Collaboration with Johns Hopkins ¢ Potential lifeline for children e POETICTrial upcoming
Slide 13: BIOLOGY IS
BIOLOGY IS
Slide 14: Data points and machine learning
of data points MACHINE LEARNING — ~ patterns)" ., >
Slide 15: LP-184, an investigational drug
LP-184 an investigational drug targeting the in cancer cells — causing double stranded breaks, and the cancer cell.
Slide 16: How could a machine reason through sparse, disconnected information?
reason through sparse, disconnected information? ~ How coul hine Why was LP-184 brilliant in solid tumors... but completely resisted in blood cancers?
Slide 17: Correlation between LP-184 and PTGR1 expression levels
@ @ Brain_Cn e @ Kidney POSITIVE between LP-184 and PTGR1 expression levels
Slide 18: LP-184 was predicted by our original AI platform (RADR)
LP-184 was predicted by our original Al platform (RADR) to be dependent on = Using RADR®, PTGR1 was Identified as a Biomarker that Predicts LP-184 Response Using RADR®, PTGR1 was Identified as a Biomarker that Predicts LP-184 Response Using RADR®, PTGR1 was Identified as a Biomarker that Predicts LP
Slide 19: RADR Precision Medicine Platform: CRISPR gene editing studies
RADR® CRISPR GENE EDITING STUDIES: ¢ No PTGR1 = no LP-184 activity ¢ Tumor cell overexpress PTGR1 e Leads to LP-184 selectively killing Precision RA | Medicine DR. Platform
Slide 20: PTGR1 correlation with LP-184 response
Pan_compound_response (mda, NCI-60) vs. PTGR1 (xai, NCI-60) (xai, NCI-60) vs. LP-184N (act, Lantern Pharma) Pearson correlation (r)=-0.55, p-value=7.9e-06 arson correlation (r)=0.88, p-value=8.4e-20 PARADOXICAL :: . @ Brain_CNS @ Lung e @ @ Ovary @, 4 @ Prostate Sve, o of @ Skin Al didn't just give
Slide 21: A Devastating Pediatric Cancer With Limited Therapeutic Options
—A Devastating Pediatric Cancer With Limited Therapeutic Options LP-184 crosses the blood-brain barrier effectively ATRT is linked to mutations in a chromatin remodeling protein called SMARCB1 Loss of SMARCB1 disrupts DNA damage repair pathways—exactly what LP-184 exploits. ATRT tumors show low expr
Slide 22: Lethal in tumors with deficiency of chromatin modeling proteins - SMARCB1
Lethal in tumors with deficiency of chromatin modeling proteins- SMARCB1- Case for ATRT SMARCB1 (INI1) and SMARCA4 are a component of the human SWI/SNF chromatin remodeling complex and found to be frequently mutated resulting loss of function in ATRTS and Rhabdoid tumors Drug sensitivity (IC50) of L
Slide 23: Subcutaneous cell line derived xenograft model
subcutaneous cell line derived xenograft model i ee. (SMARCB1 deletion, MYC elevation) ; © Vehicle LP-184 (2mg/kg) LP-184 (mg/kg) Tumor Volumes for — Tumor Volume mm? 30 35 cn Representative — terminal tumors
Slide 24: What if we built an AI specifically for...
What if we built an specifically for
Slide 25: Atypical Teratoid Rhabdoid Tumors (ATRT)
Atypical Teratoid Rhabdoid- Tumors ( ) nots ( What if we built an specifica uly for
Slide 26: Slide 26
Slide 27: Slide 27
Slide 28: Which rare cancers have the greatest unmet medical needs?
Which rare cancers have the greatest unmet medical needs? 3) Hello, User
Slide 29: Designs medicinally valid molecules
e Designs medicinally valid molecules e Several novel molecules for rare cancers in development TODAY
Slide 30: AI isn't replacing scientists
@ withZETA A.|. isn't replacing scientists. It's becoming the most colleague we've ever had.
Slide 31: Reed Bender, Senior Data Engineer, withZeta.ai Architect
ie : Reed Bender, WV Senior Data Engineer — withZeta.ai Architect — + BUILT BY @)lantempnarma withzeta.ai *
Slide 32: Zeta Architecture
ZETAARCHITECTURE @) withZETA { Derived insights are added into | conversation thread = : Zeta Toolkit Tiras cen sou Intelligence Engine D & Al Modules ‘Amazon Lambda Amazon Bedrock ® Terie Yi) ' oe ere ed ie | Publed Zea comple comprehen Nae ead core | __ research? Loops Curated Databases — t openFD
Slide 33: Recursive, Agentic Research
RECURSIVE, AGENTIC RESEARCH @) @__» Internal Chain-of-Thought Bo I need more z Initial State —————— (Cot) —> | research? ' ——+ | Tool Execution - User's Question - What is the user asking? = Asynchronous, - Mode-Specific System Prompt - How many tools have | used? serverless function call(s) - Avail
Slide 34: Zeta Toolkit
BUILT BY ZETA TOOLKIT Al Modules Zeta Toolkit Advanced machine learning ensemble that predicts blood- brain barrier permeability for small molecules with 94.1% accuracy. Uses 16 complementary ML models to evaluate molecular structures and provide CNS drug development insights for neurological therap
Lantern Pharma — Investor Presentation (January 2026) on 6ix