FirstTensor.com

THE BIOTECH GOLD RUSH IS COMING. GENESISL1 BUILDS THE SHOVELS. ⛏️ In every gold rush, the biggest fortunes were made not by those who dug, but by those who supplied the shovels and tools. The biotech gold rush is starting now. GenesisL1 runs the supply store, the first of its kind: on-chain, public, and open to every lab, every researcher and every AI agent. ━━━━━━━━━━━━ WHY NOW AI designs proteins, peptides and drugs faster than any lab can check them. Results are cheap; proof is scarce. That is the real demand of this rush: data you can verify, models you can replay, IP you can protect, and a record nobody can rewrite. ━━━━━━━━━━━━ LIVE TODAY • MolNFT: 1M+ molecular records, with 229,000+ PDB structures held in full in chain state • GL1F (GenesisL1 Forest): machine learning models as NFTs, with deterministic on-chain inference • CIPNFT: encrypted IP with post-quantum protection • gl1x.com/bio: a decentralized bioinformatics laboratory inside the explorer ━━━━━━━━━━━━ BUILT FOR AI AGENTS Much of the digging in this rush will be done by agents, not by people. GenesisL1 is equipped for them. • Tooling: 80+ tools served over MCP, 40+ of them pure bioinformatics • Sequence and structure: BLAST across 1.2M+ protein sequences, HMMER, MSA, structure search and alignment • Analysis: pockets, ligands, mutation scans, variant effects, proteomics, PubMed • Full workflow: an agent can search, build a training dataset, run an on-chain model, and prove its input data from chain state at a pinned block In development: GL1AI, an open inference market with an OpenAI-compatible API and x402 machine payments, so agents can buy compute on their own. An agent's work here can be checked, not just believed. ━━━━━━━━━━━━ PEPTL (IN DEVELOPMENT) Peptides are one of the richest veins of this rush. PEPTL is a peptide design and docking studio built on the whole stack: MolNFT for structures, GL1F and GL1AI for models. 1. Targets come straight from chain state as MolNFT objects. 2. Design and docking run as in silico experiments in the studio. 3. Results are published to the same chain, public or sealed with CIPNFT. Every experiment is preserved, so a design carries its evidence with it. ━━━━━━━━━━━━ GL1-SEQ: GIT FOR GENOMES 🧬 This month The Bioinformatics Company presents GL1-seq, with a preprint. The idea: manage biological sequence the way git manages code. Today a genome travels as one huge file. Change one base and every pipeline starts from zero. Nobody can prove which version of which sequence produced a result. How GL1-seq works: 1. Canonical encoding Every sequence has exactly one valid encoding: about 2 bits per base for DNA, with compact codes for proteins. Reconstruction is lossless and bit-exact. 2. Content-defined chunks The sequence is cut at anchors chosen by the sequence itself, not by position. Insert one base and only the adjacent chunks are re-cut. Everything else is reused as is. 3. Merkle tree Chunks are hashed into a tree, and the root is the sequence's identifier. Same sequence, same ID, on any machine, in any language, on-chain or off-chain. 4. Edits as deltas A new genome is the reference plus a delta. An edit, a slice, a duplication or a translocation rebuilds only the seams, never the whole genome. 5. Certificates Every edit and every liftover between assemblies carries a certificate sized by the change, not by the genome. Anyone holding only the original identifier can verify it, and so can the chain. Large changes are posted as bonded claims that anyone can challenge, and a dispute comes down to checking a single symbol. 6. Git-style history Commits, branches, merges and conflict reports. Annotation layers travel with the sequence, so genes are lifted from one version to the next and proteins re-derived with a verifiable receipt. 7. Partial reads with proof Read any region of any genome with proof that it belongs there, without downloading the genome. What goes into GenesisL1 chain state: • The human reference genome (GRCh38), in full • Every other genome as a delta against it: 10–20 MB per human genome • 12 pilot genomes at launch, 1000+ public genomes in the months after • The whole human GENCODE annotation, for gene-by-gene comparison and alignment • The human proteome and protein atlases: Swiss-Prot, RefSeq, UniRef50, paired antibody repertoires The result: • More than 7× faster than full recompute • Implementations in Python, JavaScript, Go and C++ that produce identical identifiers • An MCP interface for AI agents from day one Private data never goes on-chain. It stays in each institution's own compliant environment. The chain holds only cryptographic commitments: a provenance and audit layer. ━━━━━━━━━━━━ THE STORE IS OPEN Mainnet live since 2021. No founder, team or investor allocation. No token sale. The tools and supplies are public infrastructure, so any lab, any researcher and any agent can walk in and equip itself today. The rush is starting. The tools are on the shelf. ⛏️ genesisl1.com | gl1x.com/bio #GenesisL1 #DeSci #Biotech #Biotechnology #Bioinformatics #AIAgent #Peptide #TechBio #AIAgents #OnChainAI #BioFi #Web3 #BiotechGoldRush #AgenticAI #DecentralizedScience #SyntheticBiology #MachineLearning #GL1

Dear BioAI, MedAI, SynBio and biohacking community, We are very happy to present the very first on-chain bioinformatics laboratory, as a part of the broad GenesisL1 decentralized science architecture: gl1x.com/bio — Bioinformatics Onchain Free. No account. Open to humans, programs and AI agents. What it is GenesisL1 carries hundreds of thousands of protein and nucleic acid structures published as MolNFT records, stored in contract state or anchored on chain with CID-verified IPFS content. GL1X treats that archive as a scientific database and runs the standard methods of sequence and structure bioinformatics over it, in its own code, with no external services, and each method validated against its reference program. This is not a wrapper around someone else's server. It is a laboratory that lives on the chain. What you can do today 🔎 Search & homology — molecular search by PDB ID, UniProt accession, NFT id, title, sequence or motif; full BLAST (blastp/n/x, tblastn, tblastx) over every sequence, identical to NCBI BLAST+ on the test set; PROSITE-style motif scans across sequences and 3D structures; profile HMM search reproducing the hmmsearch pipeline. 🧬 Structure — TM-align superposition of any chain against any other (or your own file); structure search over every indexed chain in seconds via the GSA structural alphabet; pre-computed experiment-vs-prediction comparisons for every UniProt accession present as both. 🔬 Full-atom analysis in the research workbench — DSSP secondary structure and H-bonds, accessibility, half-sphere exposure, contact maps, base pairs with pseudoknots, chain interfaces with buried surface, salt bridges; SS-pattern search (β-hairpins, helix-turn-helix); 3D templates for catalytic triads and metal sites. 💊 Pockets & druggability — cavities with volume, buriedness, hydropathy, charge and a druggability score, searchable across every structure; a ligand index of every cofactor, drug, ion and sugar with its binding-site residues. 🧪 pH environments — structure-based pKa for every ionizable group; charge, pI and stability of a protein in the stomach, intestine, blood, cytosol, endosome or lysosome, with membrane proteins getting a pH on each side. ⚗️ Proteomics — digestion by 24 enzymes, proteotypic peptides across the archive, masses, fragments, isotope patterns, MGF spectrum identification with target-decoy FDR, crosslink checks, HDX/LiP painting, a virtual 2D gel. 🌳 Families, evolution & variants — multiple alignment at MUSCLE-level accuracy, Jensen-Shannon conservation on the structure, NJ trees, mean-field DCA coevolution vs contacts, CD-HIT clustering, disorder/coiled-coil/TM annotation, substitution scans and ΔΔG from GL1F models. 📊 Datasets & on-chain ML — turn any search into a dataset (JSON/CSV/TSV) with documented features, or straight into GL1F model inputs; train off chain, publish the model as an NFT, run it on GenesisL1 on any record. 📚 Literature — PubMed search with up to 2,000 records clustered into a topic map in your browser (Broccoli), every structure linked to its literature. ✅ Proofs — every sequence in every result can be re-derived from the chain at a pinned block and from IPFS, with the calls to replay on your own node. Every hit names the record and the SHA-256 of what it describes. Deterministic answers, reference parity, held-out evaluation. 🤖 Agent-native — every tool above is an MCP tool and an API endpoint. An agent can identify a molecule, BLAST it, find its structural neighbours, analyse its pockets, digest it, align its family, scan its variants, read its literature, build a dataset, run a model and prove the sequences, end to end. Where this is going This is the opening move. The horizon is expanding constantly: the full human reference genome and pilot genomes are coming into chain state through GL1-seq, with 1000+ public genomes to follow, alongside Swiss-Prot, RefSeq proteins, UniRef50, paired antibody repertoires and IPD-IMGT/HLA. Peptide design (PEPTL), confidential IP (CIPNFT), scientific DAOs (COLLAB) and DeSci finance (DESCIFI) are being built around the same core. Our aim is simple and ambitious: hundreds of thousands of researchers working on decentralized infrastructure, with tools and services for open science expanding at a pace the field has not seen before. This is an invitation Research on chain. Build on chain. Point your agents at the MCP server, call the API, fork the explorer, publish a model, mint your structures. Everything you compute here is provable, reproducible and belongs to no single server. → gl1x.com/bio · Docs: gl1x.com/docs/bioinformatics · API: gl1x.com/docs/api · MCP: gl1x.com/docs/mcp Welcome to the laboratory. GenesisL1

Everyone says whales are accumulating $L1. Unmask checked the chain: Whales net bought 58.9% of what they traded in 7 days. 974,893 wL1 bought vs 576,996 sold. 74% of the L1 bought in the last 24h is still held. Don't trust, verify 🔍 unmaskcrypto.io #GenesisL1

Top gainers Base coins last 24H What project are you betting big on?👇 $wL1 @genesis_L1 $MYC @MYC_TOKEN $POD @dphnAI $SLC @silencioNetwork $DGUY @StonksDGUY $PEPE @0x52BasedPepe $Fren Pet @frenpetonbase $aeon @aeonframework $RWA @RWA_Inc_ $Ratspeak @ratspeakorg

𝗚𝗲𝗻𝗼𝗺𝗲𝘀 𝗮𝗿𝗲 𝘀𝗼𝘂𝗿𝗰𝗲 𝗰𝗼𝗱𝗲. 𝗜𝘁'𝘀 𝘁𝗶𝗺𝗲 𝘄𝗲 𝗺𝗮𝗻𝗮𝗴𝗲𝗱 𝘁𝗵𝗲𝗺 𝗹𝗶𝗸𝗲 𝗶𝘁. (The git moment for genomics) This month The Bioinformatics Company unveils 𝗚𝗟𝟭-𝘀𝗲𝗾 — and the preprint behind it. One protocol that treats DNA, RNA and protein sequences the way git treats code: content-addressed, versioned, diffable, auditable. Two layers. The separation is the point. 𝗟𝗮𝘆𝗲𝗿 𝟭 — 𝗣𝘂𝗯𝗹𝗶𝗰 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲, 𝗼𝗻 𝗰𝗵𝗮𝗶𝗻 The 𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗵𝘂𝗺𝗮𝗻 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗴𝗲𝗻𝗼𝗺𝗲 (𝗚𝗥𝗖𝗵𝟯𝟴) — every base, in full, in chain state on GenesisL1. Not a hash. Not a pointer. The sequence itself, in GL1-seq's native ~2-bit encoding, 4× denser than FASTA. One canonical sequence, verifiable by anyone: the main branch the world commits against. Every other genome is a diff. A full human genome as a diff is 𝟭𝟬–𝟮𝟬 𝗠𝗕 — and thousands of genomes on chain become routine. At launch: the 𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗵𝘂𝗺𝗮𝗻 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗴𝗲𝗻𝗼𝗺𝗲 (𝗚𝗥𝗖𝗵𝟯𝟴) + 𝟭𝟮 𝗽𝗶𝗹𝗼𝘁 𝗴𝗲𝗻𝗼𝗺𝗲𝘀. Next: 𝟭,𝟬𝟬𝟬+ public genomes, then Swiss-Prot, RefSeq, UniRef50, IMGT/HLA and antibody repertoires. Not theory. GenesisL1 already holds 𝟭𝗠+ PDB and AlphaFold structures in chain state (MolNFT). Sequences are next. 𝗟𝗮𝘆𝗲𝗿 𝟮 — 𝗣𝗿𝗶𝘃𝗮𝘁𝗲 𝗱𝗮𝘁𝗮, 𝗶𝗻 𝗮 𝗰𝗼𝗺𝗽𝗹𝗶𝗮𝗻𝘁 𝗲𝗻𝘃𝗶𝗿𝗼𝗻𝗺𝗲𝗻𝘁 Patient and participant genomes 𝗻𝗲𝘃𝗲𝗿 𝘁𝗼𝘂𝗰𝗵 𝘁𝗵𝗲 𝗰𝗵𝗮𝗶𝗻. They remain in the institution's private, compliant environment — under its own governance, access controls and jurisdiction. The chain records only cryptographic commitments — a fingerprint, never the sequence. Provenance and change history, verifiable by anyone, with zero data leaving the institution. 𝗪𝗵𝗮𝘁 𝗶𝘁 𝗱𝗲𝗹𝗶𝘃𝗲𝗿𝘀 🔹 𝗟𝗼𝘀𝘀𝗹𝗲𝘀𝘀 𝗯𝘆 𝗰𝗼𝗻𝘀𝘁𝗿𝘂𝗰𝘁𝗶𝗼𝗻 — bit-exact reconstruction from reference + commits. 100% fidelity. 🔹 >𝟰× 𝗳𝗮𝘀𝘁𝗲𝗿, >𝟰× 𝗰𝗵𝗲𝗮𝗽𝗲𝗿 — touch the delta, never the full sequence. 🔹 𝗣𝗿𝗼𝘃𝗲𝗻𝗮𝗻𝗰𝗲 𝗮𝗻𝗱 𝗮𝘂𝗱𝗶𝘁 𝗯𝘂𝗶𝗹𝘁 𝗶𝗻 — who, what, derived from what. Reproducibility becomes a property of the data. 🔹 𝗗𝗲𝘁𝗲𝗿𝗺𝗶𝗻𝗶𝘀𝘁𝗶𝗰 𝗲𝘃𝗲𝗿𝘆𝘄𝗵𝗲𝗿𝗲 — identical results on-chain (EVM) and off-chain, in Python, JavaScript, Go, Rust and C++. No protocol upgrade required. 🔹 𝗡𝗮𝘁𝗶𝘃𝗲 𝘁𝗼 𝗵𝘂𝗺𝗮𝗻𝘀 𝗮𝗻𝗱 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀 — one format, one interface. Agents branch, analyze and commit with full attribution; humans audit every step. For biobanks, sequencing cores, national genome projects, pharma and regulators: one canonical reference everyone shares, plus a tamper-evident audit trail for private data — 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗺𝗼𝘃𝗶𝗻𝗴 𝗮 𝘀𝗶𝗻𝗴𝗹𝗲 𝗽𝗮𝘁𝗶𝗲𝗻𝘁 𝗴𝗲𝗻𝗼𝗺𝗲. Preprint, benchmarks and code this month — follow for the release. #Genomics #Bioinformatics #DeSci #DataProvenance #AIAgents #GenesisL1

this might be my last normal tweet before biotech goes vertical i'm not a whistleblower. I don't work at Lilly, Novo or any of the big labs but there is a massive shift happening in human biology right now. it's sitting in press releases and almost nobody reads them the last 6 months alone: - a once-daily pill roughly doubled median survival in metastatic pancreatic cancer (13.2 vs 6.6 months) - a one-time gene editing infusion cut hereditary angioedema attacks 87%, and 62% of patients stayed attack-free and off all other prevention for 6 months - a one-time gene edit for cholesterol dropped LDL up to 62%, and it held for over a year - a personalized mrna vaccine beat keytruda alone in a 1,137 patient melanoma phase 3 - retatrutide hit ~28% weight loss in a phase 3 - an ai-designed drug entered phase 3 - a pill that regrows hair passed its second phase 3 - an immunotherapy that could let some rectal cancer patients skip surgery, chemo and radiation just got fda approved - a man lived 271 days off dialysis on a pig kidney - human cells were caught passing chromosome-sized chunks of DNA to their neighbors, and it changes how the neighbors behave we are on the verge of an era where the human body is a software problem and the update cycle gets shorter everyday stick around, it's going to get weird bio/acc