Taichi Matsubara, Ph.D.

Ph.D. in Science | Researcher @ BlueMeme Inc. & Collaborative Researcher @ Kyushu University | Genomics / BioInfor / QuantumMachineLearning

[Taichi Matsubara, Ph.D. @TaichiMResearch]

Thrilled to announce that my first 1st-author paper has been published in Briefings in Bioinformatics! 🎉🧬
@OUPAcademic

​Introducing QTFPred: A robust quantum-classical hybrid model that predicts transcription factor binding at base resolution. 🧬⚛️
​📄 Paper: https://t.co/iNyAaPVBxQ
​
​#Bioinformatics #QuantumComputing #Quantum #Genomics

[Andreas Gschwind @argschwind]

Our paper is now out in Nature! 🎉

Expanded to 92M+ enhancer–gene links across 1,458 biosamples, with new benchmarks and analyses.

"An encyclopedia of human enhancer–gene regulatory interactions":
https://t.co/zOWJu64wrp

Grateful to the whole team of this massive effort!

[Ekin Deniz Aksu (ekindea.bluesky) @ekinda]

I'm excited to share that Corgi is out in @NatureComms!

We refine context-aware sequence-to-function models by using FiLM to condition the model on trans-regulator expression levels. This way Corgi can generalize to unseen cell types.

https://t.co/R2YXlpIx1L

🧵Thread 1/n

[Taichi Matsubara, Ph.D. @TaichiMResearch]

original
https://t.co/j8eBROYGO5

[Taichi Matsubara, Ph.D. @TaichiMResearch]

AI coding agents often follow your rules for a few turns—then quietly drift.

I open-sourced a simple fix: one block for CLAUDE.md or AGENTS.md that re-anchors the rules every turn and asks for y/n before acting.

EN/JA/ZH, MIT. English tested on Claude Code and Codex.

https://t.co/vs83hTzdpD

[PRX Quantum @PRX_Quantum]

On the cover: A study that advances practical #QuantumComputing by enabling a partially fault-tolerant molecular energy computation on a trapped-ion quantum computer through noise-aware calibration and quantum error-correction techniques.

🔗 https://t.co/nb8j77vCRq

[PRX Quantum @PRX_Quantum]

Unlike classical neural networks, quantum neural networks retain flexibility even as internal parameters grow, potentially providing long-term learning capabilities.

Read the paper: https://t.co/mdqgkPRBM3

[Anthropic @AnthropicAI]

We’ve received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5.

We'll begin restoring access tomorrow, and will share an update soon.

We’re grateful to our users for their patience, and to everyone who worked with us on redeploying the models.

[Claude @claudeai]

Introducing Claude Science, a new app designed with every stage of research in mind.

Artifacts traced to their code, environments managed on demand, and 60+ optional scientific databases that you can connect.

Available now in beta.

[Taichi Matsubara, Ph.D. @TaichiMResearch]

https://t.co/b12GXmjgkp

[Taichi Matsubara, Ph.D. @TaichiMResearch]

Introduction to quantum computation and quantum machine learning.

And these are related to qtfpred https://t.co/gHZkUfwwqM , which is quantum-classical-hybrid tf-dna-binding predictor

[Taichi Matsubara, Ph.D. @TaichiMResearch]

this is a quantum introduction codes for bioinformatics
https://t.co/2Nn2XJcwK6

[生化学若い研究者の会（生化学若手の会） @seikawakate]

九州支部です🌟
大分で開催される生化学会九州支部例会にて、「Hands-onで学ぶ 研究に活かせるバイオインフォマティクス」を開催します🧬
松原太一博士@TaichiMResearch (株式会社BlueMeme)と、S. Alireza KHANGHAHIさん(九大)をお呼びし、生命情報解析の基礎から応用までをご紹介いただきます。(1/2)

[Taichi Matsubara, Ph.D. @TaichiMResearch]

I'm excited to speak at the Hands-on Bioinformatics Symposium organized by @seikawakate Kyushu Branch this Sunday, June 28 in Oita!

[たにぐち まこと／ちゃんとWeb withAI @seltzer]

マネフォさんは、もうダメかもしれないなー。法人の会計、乗り換えるかどうしよう、、

マネーフォワード、約6．3万人分の情報流出(時事通信)
#Yahooニュース
https://t.co/daVIDQ7ESQ

[Taichi Matsubara, Ph.D. @TaichiMResearch]

QTFPred is publicly available on Github. You can check it out here:
https://t.co/f4hINqg3nT

[すみか@Skillsオタク @sumika45379]

【結論から言う】
Claude Codeで論文を書く人は、Academic Research Skillsを入れたかどうかで別物だよね😳
海外で伸びてたのは
30以上の専門agentをまとめた、研究チーム型のClaude Skill。
論文をraw ideaからpublication-readyまで持っていきたい人は、まず見てほしい。

① Deep research
arXiv / Semantic Scholar / DBLPを横断して調べる。
「まず関連研究を集める」がSkill化されてるの普通に強いんだけど、、

② Drafting
paperをsection by sectionで作る。
一気に全部書かせるんじゃなくて、
論文の構成に沿って進めるやつ。

③ Integrity check
referenceとdataを100% verificationする前提。
ここが甘いAI論文作成と違う。

④ Peer review
5-person peer reviewにDevil's Advocateまで入る。
自分に甘い文章をちゃんと壊しにくるのが良いんだよね。

⑤ Export
APA / IEEE / Chicago / MLA / Vancouverに対応。
最後の形式合わせまでSkill側に寄せられる。

しかも導入は
/plugin install academic-research-skills
これだけ。

ここまで入れると👇
「Claudeに論文を手伝わせる」
じゃなくて
「Claude Codeに研究チームを持たせる」
に変わるんだよね。

このレベルのSkill、研究やってる人には普通に神すぎる😻

研究チームSkillを使うなら、Claude Skills全体の考え方も一緒に見ると理解が深まるやつ。
この下の記事を読むと理解が一気に深まる。マジでおすすめ👇😻

[Taichi Matsubara, Ph.D. @TaichiMResearch]

https://t.co/gM0OdRQIDX

[Peter Koo @pkoo562]

[SAVE THE DATE] MLCB 2026 will be held Nov 16–17 at the New York Genome Center in NYC!

Submission deadline is July 1 AOE
• 8-page full papers: oral or poster, with optional proceedings
• 2-page work-in-progress papers: posters

Spread the word! RT!

https://t.co/BTmQMzLN4d

[高遠 頼@生命科学VTuber(たかとー らい) 定期科学ニュース配信(日)🧬🔬 🥼 @takatoh_life]

博士課程に来なさい

[はかせチャン @hshimodaira]

ML/NLPトップ会議にバンバン通せる層は、ギュられる恐怖で焦って修士就職するよりD進した方がプラスかもよ？

論文は完全に個人資産だから、社会がギュられて転職・キャリアチェンジが普通になる時代でも価値が消えない。
数年後にまたキャリアアップが必要になるなら、最初から最先端で研究続けた方がええんちゃう？しらんけどw