Tag: Academic research

Posted on

StockWatch: Trump Order Lifts Psychedelic Drug Shares

Stocks of most publicly traded psychedelic drug developers jumped when President Donald Trump signed Executive Order 14401, directing the FDA and other federal agencies to accelerate research and improve access to psychedelic drugs, citing their potential as promising treatments for serious mental illnesses.

Among its provisions, the order directs the FDA to provide Commissioner’s National Priority Vouchers (CNPVs) to “appropriate” psychedelic drugs that were granted the agency’s Breakthrough Therapy designation and met the voucher program’s criteria. The FDA’s parent agency, the Department of Health and Human Services (HHS), is required to spend at least $50 million through the Advanced Research Projects Agency for Health (ARPA-H) “to support and partner with” state governments that have enacted or are developing programs to advance psychedelic drugs for serious mental illnesses.

“This is an unmet public health need and potentially promising treatments. That’s why there’s a sense of urgency around this, and why we’re doing it now,” FDA Commissioner Martin A. Makary, MD, said at the ceremony where Trump signed the order. “Applications are about to come in, and this is the perfect timing for this announcement.”

At least one analyst agreed that the timing was right for Washington to spur the development of psychedelic drugs.

“Investor mindshare should rise meaningfully ahead of pot’l approvals in 2027–30,” Andrew Tsai, equity analyst with Jefferies, observed in a research note. “As we approach the first pot’l FDA approval of a psychedelic in 2027, President Trump is providing an official stamp of validation to the class in the form of an executive order, reassuring us that the FDA/HHS/White House’s support of psychedelics is real/actionable (not rhetorical).”

Proving correct

Makary said the FDA planned to issue CNPVs to three serotonin 2a agonists, a class that includes LSD and other psychedelic drugs. While he did not reveal specific companies and drugs by name, market watchers immediately speculated that one of the drugs was COMP360 synthetic psilocybin, the lead clinical candidate of Compass Pathways (NASDAQ: CMPS)—speculation that proved correct when COMP360 won a CNPV on Friday.

COMP360 is expected, according to Tsai, to be the first psychedelic drug to win FDA approval in 2027. In February, Compass announced what it called statistically significant and clinically meaningful data from two Phase III trials assessing COMP360 in treatment-resistant depression (TRD), COMP005 (NCT05624268) and COMP006 (NCT05711940). The data showed positive effects for COMP360 within one day, lasting at least through six months after just one or two doses among those who have a clinically meaningful response.

COMP360 is also in Phase II trials for both PTSD and anorexia nervosa.

Compass fueled speculation about an FDA voucher approval by issuing a statement supporting the executive order: “Today’s announcement aligns regulatory urgency with patient need, and we applaud the Administration for taking this important step forward in accelerating access, without compromising rigorous science.”

Investors celebrated with Compass, whose shares soared 42% from $6.66 to $9.46 on April 20, the first trading day after the order signing. Shares yo-yoed the rest of the week, sliding 7.5% to $8.75 Wednesday before rebounding nearly 5% to $9.15 Thursday and rising another roughly 5% to $9.58 Friday on news of the voucher approval. Year-over-year, Compass shares have more than doubled, soaring about 140% from $5.22 on April 24, 2025.

FDA names additional voucher grantees

The FDA indeed issued three CNPVs on Friday—one to Compass as previously mentioned, one to Usona Institute, a nonprofit medical research organization, for psilocybin for major depressive disorder (MDD), and one to Otsuka Pharmaceutical (Tokyo Stock Exchange: 4578) for methylone (TSND-201) for post-traumatic stress disorder (PTSD). Otsuka is acquiring the methylone program as part of its up-to-$1.225 billion ($700 million upfront) purchase of privately held Transcend Therapeutics, announced last month.

Launched in October by Makary, CNPVs are awarded to drug developers whose work is deemed to address a health crisis in the United States, deliver more innovative cures, address unmet public health needs, and increase domestic drug manufacturing as a national security issue. The vouchers entitle companies to reviews of their final applications within a target timeframe of 1–2 months, rather than the current 10–12 months.

“Ultimately, we do not see the FDA’s issuance of the first set of CNPVs as precluding other psychedelic players from also obtaining CNPVs in the future—so we think the FDA’s action today bodes well for the space broadly,” Tsai wrote after the FDA announced the voucher recipients. “Net-net, the macro backdrop for psychedelics is improving.”

That improvement, Tsai added, reflects Trump’s endorsement of psychedelic drugs, a collaborative FDA, and growing interest in the space by big pharma giants such as Johnson & Johnson (NYSE: JNJ), which generated $1.696 billion in 2025 sales and $468 million in first quarter sales from Spravato (esketamine), an NMDA receptor antagonist indicated for treatment-resistent depression (TRD) and depressive symptoms in adults with MDD with acute suicidal ideation or behavior in conjunction with an oral antidepressant.

The voucher decision hardly budged Otsuka shares, which dipped nearly 1% Friday from ¥10,870 ($68.18) to ¥10,810 ($67.80).

However, Compass was one of several psychedelic drug companies to see their shares surge on news of the executive order.

AtaiBeckley (NASDAQ: ATAI), formed last November by the merger of atai Life Sciences and Beckley Psytech, jumped 22% from $4.03 to $4.90 on April 20, then plateaued the rest of the week, finishing Friday at $4.63 and a 15% one-week gain. AtaiBeckley shares year-over-year have more than tripled, rocketing 204% from $1.53 a year ago Friday.

Definium Therapeutics (NASDAQ: DFTX) shares rose 5% over two days, from $22.68 the Friday before Trump signed the order to $23.84 on Tuesday, but gave back all the week’s gain, finishing Friday at $22.48. Long-range investors have fared better, as Definium shares have more than tripled, zooming 249% from $6.43 on April 24, 2025.

GH Research (NASDAQ: GHRS) shares climbed 17% from $18.34 to $21.50 the first day after the executive order, only to drop 6% the rest of the week, closing Friday at $20.25 and settling for a 10% one-week gain. GH’s shares doubled year-over-year, growing 101% from $9.50 a year ago Friday.

Showing volatility

The executive order wasn’t enough to boost shares of Cybin, which operates under the name Helus Pharma (NASDAQ: HELP). Helus showed the most volatility in the days following the order signing, tumbling 12% from $5.61 to $4.93 on April 20. The drop followed Helus’ announcement that its CEO, Michael Cola, stepped down immediately at the request of its board, succeeded by interim CEO Eric So, while the board carries out a search for a permanent chief executive.

“It’s like they can’t give investors a break,” fumed “SamZaki320” on a Reddit chat board. “The only positive is that the executive order sentiment is pushing back against a total disaster. Not that it’s a good thing, other companies are up double digits.”

Helus shares bounced back 17% to $5.77 Wednesday and dipping 0.35% to $5.75 Thursday despite the company announcing two powerhouse additions to its scientific advisory board—Robert Langer, ScD, the David H. Koch Institute professor at MIT and a co-founder of Moderna (NASDAQ: MRNA); and Stephen Brannan, MD, a neuroscience drug development expert with over 20 years of experience designing and implementing clinical programs for psychiatric and neurological disorders. Shares fell 2% Friday, closing at $5.61.

In a statement, interim CEO So lauded Trump’s order: “The Executive Order reflects growing recognition of the urgent need for new treatment options in serious mental health conditions and the importance of advancing innovative therapies through rigorous, research-based development.”

Looking beyond Washington

Yet So acknowledged that Washington alone can’t advance psych drug development beyond what its science can accomplish: “Policy momentum is meaningful, but the future of this field will ultimately be determined by the strength of the clinical evidence and the ability to deliver safe, reliable treatments at scale.”

As did Helus and Compass, Definium also praised the executive order: “We applaud the Administration’s recognition that psychedelic medicines may represent meaningful new treatment options for patients,” Definium CEO Rob Barrow stated. He cited his company’s clinical development program for DT120 (lysergide tartrate) for conditions that include generalized anxiety disorder (GAD) and MDD.

At the ceremony where he signed the executive order, Trump acknowledged being asked to address psych drug development by podcaster Joe Rogan and others, which the president said led to talks with Makary as well as HHS Secretary Robert F. Kennedy Jr., NIH Director Jay Bhattacharya, MD, PhD, and Mehmet Oz, MD, administrator for the Centers for Medicare & Medicaid Services.

“Research has been going on for quite some time. But usually with things like this, nothing ever happens, no matter how the research ends up. We’re changing that,” Trump said. “Why would we wait three or four years to get it done? Or 10 years? Frankly, let’s get it done immediately—and that’s what happened.”

Leaders and laggards

  • Daiichi Sankyo (Tokyo Stock Exchange: 4568) shares slipped 10% from ¥2,790 ($17.50) to ¥2,499 ($15.67) on Friday after the drug developer announced it was delaying the release of its annual earnings results for the fiscal year that ended March 31, from April 27 to May 11,  “as additional time is required to finalize the financial figures.” May 11 is the day when Daiichi Sankyo plans to release its five-year business plan. “The company is currently reviewing the supply plans for its oncology products portfolio and development pipeline in light of rapidly changing business conditions. As a result, additional deliberation is required to reasonably estimate the amount of loss provisions to be recorded in connection with contracts with contract manufacturers,” Daiichi Sankyo added in a statement.
  • Inhibrx Biosciences (NASDAQ: INBX) shares leaped 37% from $84.08 to $115.09 Wednesday after Reuters reported, citing unnamed sources, that Merck & Co. (NYSE: MRK), Merck KGaA (XETRA: MRK), and Ono Pharmaceutical (Tokyo Stock Exchange: 4528) were in talks with Inhibrx for a joint spinoff of two precision-engineered cancer candidates, INBRX-106 and ozekibart (INBRX-109). The treatments could have a combined value of more than $9 billion if their clinical trials prove successful, the report stated. Inhibrx declined to comment, while the other companies cited did not respond to Reuters queries. INBRX-106 is a hexavalent sdAb-based, OX40-targeting candidate being studied as monotherapy and in combination with Merck & Co.’s cancer immunotherapy blockbuster Keytruda® (pembrolizumab). Ozekibart is a tetravalent death receptor 5 (DR5) agonist antibody designed to exploit the tumor-biased cell death induced by DR5 activation. On Tuesday, Inhibrx announced ozekibart showed positive data in a Phase I/II trial (NCT03715933) assessing the drug plus Folfiri in patients with locally advanced or metastatic, unresectable colorectal cancer.
  • Organon (NYSE: OGN) shares surged 31% from $8.60 to $11.26 Friday after the Indian news outlet The Economic Times reported that Sun Pharmaceutical Industries (NSE: SUNPHARMA and BSE: 524715) had submitted a $13 billion offer for the women’s health drug developer spun out of Merck & Co. (NYSE: MRK) in 2021. The deal would be Sun’s largest ever merger-and-acquisition (M&A) deal—if Sun can prevail over at least two other would-be suitors for Organon, German-based private family-owned drug developer Grünenthal, and EQT (Nasdaq Stockholm: EQT), a Swedish-based global investment organization. The latest surge comes two weeks after Organo shares zoomed 28% on an April 10 Economic Times report stating that Sun Pharma had submitted a $12 billion all-cash offer for Organon.
  • Spruce Biosciences (NASDAQ: SPRB) shares tumbled 26% from $69.89 to $51.69 Tuesday after the neurological disorder drug developer priced a $69 million offering of common stock and pre-funded warrants that generated $64.4 million in net proceeds. The offering consisted of 1.15 million shares of common stock priced at $50 per share and pre-funded warrants to purchase 50,000 shares at $49.99 per share. Al shares and pre-funded warrants were sold, and underwriters of the offering exercised in full their option to purchase up to an additional 180,000 shares at the public offering price on Tuesday. “We intend to use the net proceeds from this offering to advance the company’s pre-commercial and launch activities, for planned clinical trials, and for working capital, capital expenditures, and other general corporate purposes,” Spruce stated in its prospectus supplement filed Tuesday. Leerink Partners, Guggenheim Securities, and Oppenheimer & Co. acted as joint book-running managers, while Jones and Craig-Hallum acted as co-managers for the offering.

The post StockWatch: Trump Order Lifts Psychedelic Drug Shares appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

Medication Treatment for Tics and Tourette’s

There are several kinds of medication than can help kids with Tourette’s or another tic disorder. But it’s important to note that not all kids who develop tics need treatment. Tics are very common. They often go away on their own, and they tend to bother parents more than they do the children experiencing them. Drawing attention to them can make them worse. So doing nothing can be the best strategy — at least initially.

Treatment comes into play if tics are upsetting your child, giving them pain, or making it hard for them to function in everyday life — say they’re disrupting class or getting bullied because of their tics.

The first recommended step in treatment is a specialized form of therapy called comprehensive behavioral intervention for tics (CBIT). CBIT is centered on habit reversal training, in which the child learns to recognize when they have an urge to tic and substitute a competing response — an easier, more comfortable, or less noticeable action or behavior that makes the tic impossible. For instance, if a child’s tic is jerking their head to the side, the strategy might be to put their chin down instead.

But if therapy isn’t effective in reducing a child’s tics, medication can help.

Guanfacine and clonidine for tics

First-line medications for Tourette’s and other tic disorders are a class of drugs called alpha-2 agonists, explains Paul Mitrani, MD, PhD, a child and adolescent psychiatrist at the Child Mind Institute. Alpha agonists decrease the release of a neurotransmitter called norepinephrine, which stimulates the nervous system. Alpha agonists serve as a kind of dimmer switch — by calming down the system, they make the urge to tic less frequent, less intense, and by extension, easier to control.

The two alpha-2 agonists usually prescribed for tics are guanfacine and clonidine. Dr. Mitrani reports that he usually starts by prescribing guanfacine because it comes in a longer-acting form (Intuniv), which reduces symptoms for a full 24 hours. Clonidine’s long-acting form (Kapvay) is effective for 12 hours.

Dr. Mitrani adds that there is a new liquid form of clonidine called Onyda XR that lasts 24 hours, but there isn’t yet a strong body of evidence regarding its effectiveness for tics. Onyda XR is FDA-approved for ADHD, as are Kapvay and Intuniv.

While no alpha agonist medications are FDA-approved specifically for tics, Kapvay and Intuniv are frequently used off-label for them. There is ample research on their effectiveness for tics, and they are recommended by clinical practice guidelines.

Some children respond better to several doses of short-acting guanfacine or clonidine, Dr. Mitrani notes, rather than a smoother dose of a long-acting medication. This may be because medication can be timed to peak at times when kids need tic suppression most, such as at school.

Alpha agonists are the preferred first line medications for tic disorders because their side-effects, including drowsiness and low blood pressure, are relatively mild.

Antipsychotics for tics

If alpha agonists aren’t helping, the next step would be to try an antipsychotic medication, which can be more effective for treating tics, Dr. Mitrani notes, but their side effects are potentially more difficult to tolerate.

Aripiprazole (Abilify), which is FDA-approved for tics, is often Dr. Mitrani’s first choice among the antipsychotic medications. Abilify is a second-generation, or atypical, antipsychotic, a group of medications that have fewer side effects than older antipsychotics. Side effects of Abilify can include restlessness, agitation and weight gain.

Haloperidol (Haldol) is also effective for tics, but it’s an older antipsychotic with more side effect concerns, Dr. Mitrani notes. “I’ve only had one patient ever on Haldol, and he tolerated it well and it really helped with his tics when other things did not.”

Risperidone (Risperdal) is another atypical antipsychotic that can help, but its side effects tend to be worse than Abilify. Risperidone can cause more concerning weight gain and metabolic, neurological, and hormonal changes that can be harmful. Sometimes other medications are used to manage the weight gain from antipsychotics.

When kids with tics also have ADHD

More than three-quarters of kids diagnosed with a tic disorder also have another disorder. When a child has multiple disorders, a clinician will want to evaluate which is causing the child the most difficulty and prioritize treating that.

The most common co-occurring disorder with tics is ADHD. “If tics are the bigger problem, we would start with treating them,” says Dr. Mitrani. “If the ADHD is the bigger problem, which it typically is, we usually treat that first.”

In the past, it was recommended that children with tics and ADHD avoid stimulant medication, based on research that showed it made tics worse. But newer studies counter that finding, Dr. Mitrani notes, concluding that the old research was based on very high doses of amphetamine-based medications. To lower the risk of exacerbating tics, he recommends starting kids with ADHD and tics on methylphenidate-based medication.

“If your child is starting a stimulant,” he adds, “and you see worsening of tics — and it’s clearly related to when the stimulant is in their system — the best approach might be a lower dose of stimulant combined with guanfacine or clonidine.”

One advantage to that combination, he notes, is that kids with ADHD who have behavior problems can benefit from the guanfacine or clonidine being active in the mornings before the stimulant starts working and in the evenings when it’s out of their system.

Kids with other co-occurring disorders

When children with tics have other co-occurring disorders, such as anxiety, OCD, or depression, treating them with medication needs to be done very carefully, Dr. Mitrani says. Since children are typically not bothered by the tics themselves, it’s almost always the other disorder that is more problematic for them.  And, he adds, when the other problems cause distress, it can make the tics worse.

For anxiety, OCD, and depression, the first-line medication treatment is an antidepressant. Antidepressants can actually help alleviate tics indirectly, since they reduce anxiety. “Stress increases tics, so if there is significant anxiety and you treat the anxiety, the tics may get better,” Dr. Mitrani says. “And then maybe you don’t need the guanfacine or clonidine. But again, it depends on what the co-occurring disorders are and what’s the bigger problem for the child.”

Monitoring medication for tics

Due to the waxing and waning nature of tics, it can be challenging to see the full effect of medication and other interventions. It is important to give medication enough time to work, Dr. Mitrani notes, typically a few weeks, to see if the overall pattern, frequency, and severity of tics has improved. And children who are being treated should continue to be monitored regularly for any changes, as tics can recur or worsen, especially when a child is excited, tired, or experiencing more stress.

Most children with tics see a natural improvement or even resolution of tics as they progress through adolescence. If there seems to be a long-standing improvement, it is appropriate to consider reducing or stopping medication, especially if the child is experiencing side effects, Dr. Mitrani notes. If tics continue and are causing distress, it is important to keep treating them.

A child going off any of these medications — alpha agonists or antipsychotics — should do so gradually, by having their dose reduced over weeks or even longer, to avoid unpleasant or dangerous side effects of sudden withdrawal.

The post Medication Treatment for Tics and Tourette’s appeared first on Child Mind Institute.

Posted on

Nancy Cox, a CDC veteran and a stalwart in global flu research, dies at 77

Nancy Cox, who for decades was a global leader in influenza research, has died. Cox headed the influenza team at the Centers for Disease Control and Prevention for 22 years, shepherding it from a branch of 14 people to a division of over 100. She was also director of the World Health Organization’s Collaborating Center for the Surveillance, Epidemiology, and Control of Influenza at the CDC.

Cox died Thursday from glioblastoma, a cancer of the brain. She was 77.

Read the rest…

Posted on

Associations of psychological distress, gaming motives and internet gaming disorder in adolescents: a network analysis

Background and objectiveThe rapid popularization of the Internet among Chinese adolescents has resulted in the emergence of a public major concern known as Internet Gaming Disorder (IGD). As demonstrated by previous studies, an association has been demonstrated among emotional distress, gaming motives and IGD. Nevertheless, the specific pathways connecting these constructs remain to be elucidated. The present study aims to explore the network structure characterizing the interactions among these three constructs and to identify potential targets for psychological interventions.MethodsThis was a cross-sectional survey conducted in city of Hangzhou. A total of 3,795 middle school students were included in the analysis. The 21-item Depression Anxiety Stress Scale (DASS-21), the Motives for Online Gaming Questionnaire (MOGQ), and the Chinese version of the Ten-Item Internet Gaming Disorder Test (IGDT-10) were used to assess emotional distress, gaming motives and IGD symptoms, respectively. Network analyses were performed using R4.5.1 software to explore the interrelationships among emotional distress, gaming motives and IGD symptoms, and identify the core symptoms and bridge symptoms.ResultsIn the depression combined network model, the presence of bridge symptoms was indicated by no initiative (D2), gaming for escape or mood relief (IGD8) and fantasy motive (fan). In anxiety combined network model, the bridge symptoms included coping motive(cop), gaming for escape or mood relief (IGD8), withdrawal (IGD2), mouth dryness (A1), and fear of embarrassment (A4). The bridge symptoms in the stress combined network model were gaming for escape or mood relief (IGD8), difficulty winding down (S1), withdrawal (IGD2), nervous energy expenditure (S3), and coping motive (cop).ConclusionThe present study explored complex network structure among psychological distress, gaming motivation, and IGD. and suggested fantasy and coping motive as bridges connecting psychological distress and IGD. Besides, our research identified no initiative, mouth dryness, difficulty winding down, fear of embarrassment, and nervous energy expenditure as the best targets for intervention to reduce IGD.
Posted on

Three reasons why DeepSeek’s new model matters

On Friday, Chinese AI firm DeepSeek released a preview of V4, its long-awaited new flagship model. Notably, the model can process much longer prompts than its last generation, thanks to a new design that helps it handle large amounts of text more efficiently. Like DeepSeek’s previous models, V4 is open source, meaning it is available for anyone to download, use, and modify.

V4 marks DeepSeek’s most significant release since R1, the reasoning model it launched in January 2025. R1, which was trained on limited computing resources, stunned the global AI industry with its strong performance and efficiency, turning DeepSeek from a little-known research team into China’s best-known AI company almost overnight. It also helped set off a wave of open-weight model releases from other Chinese AI firms. 

DeepSeek has kept a relatively low profile since then—but earlier this month, it effectively teased V4’s release when it added “expert” and “flash” modes to the online version of its model, prompting speculation that the updates were tied to a bigger upcoming release.

While the company has become a powerful symbol of China’s AI ambitions, its big return to cutting-edge frontier models comes after months of scrutiny—including major personnel departures, delays to previous model launches, and growing scrutiny from both the US and Chinese governments. 

So, will V4 shake the AI field the way R1 did? Almost certainly not, but here are three big reasons why this release matters.

1. It breaks new ground for an open-source model.

As with R1 before it, DeepSeek claims that V4’s performance rivals the best models available at a fraction of the price. This is great news for developers and for companies using the tech, because it means they can access frontier AI capabilities on their own terms, and without worrying about skyrocketing costs.

The new model comes in two versions, both of which are available on DeepSeek’s website and in its app, with API access also open to developers. V4-Pro is a larger model built for coding and complex agent tasks, and V4-Flash is a smaller version designed to be faster and cheaper to run. Both versions offer reasoning modes, in which the model can carefully parse a user’s prompt and show each step as it works through the problem.

For V4-Pro, DeepSeek charges $1.74 per million input tokens and $3.48 per million output tokens, a fraction of the cost of comparable models from OpenAI and Anthropic. V4-Flash is even cheaper, at about $0.14 per million input tokens and about $0.28 per million output tokens, making it one of the cheapest top-tier models available. This would make it a very appealing model to build applications on.

In terms of performance, V4 is, perhaps unsurprisingly, a huge jump from R1—and it seems to be a strong alternative to just about all the latest big AI models. On the major benchmarks, according to results shared by the company, DeepSeek V4-Pro competes with leading closed-source models, matching the performance of Anthropic’s Claude-Opus-4.6, OpenAI’s GPT-5.4, and Google’s Gemini-3.1. And compared to other open-source models, such as Alibaba’s Qwen-3.5 or Z.ai’s GLM-5.1, DeepSeek V4 exceeds them all on coding, math, and STEM problems, making it one of the strongest open-source models ever released. 

DeepSeek also says that V4-Pro now ranks among the strongest open-source models on benchmarks for agentic coding tasks and performs well on other tests that measure ability to carry out multistep problems. Its writing ability and world knowledge also leads the field, according to benchmarking results shared by the company. 

In a technical report released alongside the model, DeepSeek shared results from an internal survey of 85 experienced developers: More than 90% included V4-Pro among their top model choices for coding tasks.

DeepSeek says it has specifically optimized V4 for popular agent frameworks such as Claude Code, OpenClaw, and CodeBuddy.

2. It delivers on a new approach to memory efficiency.

One of the key innovations of V4 is its long context window—the amount of text the model can process at once. Both versions can handle 1 million tokens, which is large enough to fit all three volumes of The Lord of the Rings and The Hobbit combined. The company says this context window size is now the default across all DeepSeek services and it matches what is offered by cutting-edge versions of models like Gemini and Claude. 

But it’s important to know not just that DeepSeek has made this leap, but how it did so. V4 makes significant architectural changes to the company’s former models—especially in the attention mechanism, which is the feature of AI models that helps them understand each part of a prompt in relation to the rest. As the prompt text gets longer, these comparisons become much more costly, making attention one of the main bottlenecks for long-context models.

DeepSeek’s innovation was to make the model more selective about what it pays attention to. Instead of treating all earlier text as equally important, V4 compresses older information and focuses on the parts most likely to matter in the present moment, while still keeping nearby text in full so it does not miss important details. 

DeepSeek says this sharply reduces the cost of using long context. In a 1-million-token context, V4-Pro uses only 27% of the computing power required by its previous model, V3.2, while cutting memory use to 10%. The reduction in V4-Flash is even larger, using just 10% of the computing power and 7% of the memory. In practice, this could make it cheaper to build tools that need to work across huge amounts of material, such as an AI coding assistant that can read an entire codebase or a research agent that can analyze a long archive of documents without constantly forgetting what came before.

DeepSeek’s interest in long context windows didn’t start with V4. Over the past year and a half, the company has quietly published a series of papers on how AI models “remember” information, experimenting with compression and mathematical techniques to extend what AI models could realistically handle.

3. It marks the first steps on the hard road away from Nvidia.

V4 is DeepSeek’s first model optimized for domestic Chinese chips, such as Huawei’s Ascend—a move that has turned the launch into something of a test of whether China’s homegrown AI industry can begin to loosen its dependence on US chip giant Nvidia. 

This was largely expected, since The Information reported earlier this month that DeepSeek did not give American chipmakers like Nvidia and AMD early access to V4, though prerelease access is common to allow chipmakers to optimize support of the new model ahead of a launch. Instead, the company reportedly gave early access only to Chinese chipmakers. 

On Friday, Huawei said its Ascend supernode products, based on the Ascend 950 series, would support DeepSeek V4. This means that companies and individuals who want to run their own modified version of Deepseek V4 will be able to use Huawei chips easily.

Reuters previously reported that Chinese government officials recommended that DeepSeek integrate Huawei chips in its training process. And this pressure fits a broader pattern in China’s industrial policy: Strategic sectors are often pushed, and sometimes effectively required, to align with national self-reliance goals. But there’s a particular urgency when it comes to AI. Since 2022, US export controls have cut Chinese firms off from Nvidia’s most powerful chips, and they later also restricted access to downgraded China-market versions. Beijing’s response has been to accelerate the push for a domestic AI stack, from chips to software frameworks to data centers.

Chinese authorities have reportedly been pushing data centers and public computing projects to use more domestic chips, including through reported bans on foreign-made chips, sourcing quotas, and requirements to pair Nvidia chips with Chinese alternatives from companies such as Huawei and Cambricon. 

Still, replacing Nvidia is not as simple as swapping one chip for another. Nvidia’s advantage lies not only in its chips, but in the software ecosystem developers have spent years building around them. Moving to Huawei’s Ascend chips means adapting model code, rebuilding tools, and proving that systems built around those chips are stable enough for serious use.

To be clear, DeepSeek does not appear to have fully moved beyond Nvidia. The company’s technical report reveals that it is using Chinese chips to run the model for inference, or when someone asks the model to complete a task. But Liu Zhiyuan, a computer science professor at Tsinghua University, told MIT Technology Review that DeepSeek appears to have adapted only part of V4’s training process for Chinese chips. The report does not say whether some key long-context features were adapted to domestic chips, so Liu says V4 may still have been trained mainly on Nvidia chips. Multiple sources who spoke on the condition of anonymity, due to political sensitivity around these issues, told MIT Technology Review that Chinese chips still don’t perform as well as Nvidia chips but are better suited for inference than training.

DeepSeek is also tying the future costs of V4 to this hardware shift. The company says V4-Pro prices could fall significantly after Huawei’s Ascend 950 supernodes begin shipping at scale in the second half of this year. 

If that works, V4 could be an early sign that China is successfully building a parallel AI infrastructure.

Posted on

One Biosciences Chooses Albany, NY, as Its U.S. Location

Paris-based One Biosciences, an Institut Curie-backed startup, plans to set up, staff, and equip a high-complexity lab and computational analytics operation in Albany, NY, as its first U.S. location.

Empire State Development is supporting this expansion with up to $525,000 in performance-based Excelsior Jobs Program tax credits in exchange for the company’s job commitments, which anticipate 42 life science jobs and $18 million in investments over the next five years.

Officials at One Biosciences say the company will bring its proprietary technology to the first-of-its-kind hub in Albany to address the unmet clinical and scientific needs to characterize the tumor ecosystem by means of a single-cell profiling approach.

“We are excited to accelerate support of our pharma, biotech, and academic collaborators through our AI-driven single-cell technologies, which will ultimately benefit physicians and their patients,” added Vincent Miller, MD, executive chairman, One Biosciences. “The local Albany life sciences ecosystem gives us access to a community of like-minded researchers and physicians committed to leveraging technology to improve health and is an ideal location from where to serve the U.S. globally.”

“Life science research and development is vital to creating the treatments that help people heal, survive and live longer,” said New York governor Kathy Hochul. “Through our targeted efforts, we are working to ensure that cutting edge companies like One Biosciences not only grow here, but that the next generation of medical breakthroughs happen in New York State.”

The post One Biosciences Chooses Albany, NY, as Its U.S. Location appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

New Low-Toxicity Transplant Method Reverses Type 1 Diabetes in Mice

Researchers at Stanford Medicine say they have developed a combination treatment method that cured or prevented type 1 diabetes in mouse models by pairing blood stem cell transplantation with pancreatic islet cell transplantation under a substantially reduced preconditioning regimen. The approach creates a mixed immune system from both donor and recipient cells, which stopped autoimmune destruction of insulin-producing cells while also producing long-term tolerance to the transplanted tissue. The findings, published in the Journal of Clinical Investigation Insight, show that reversing type 1 diabetes can be accomplished without chronic immunosuppression or the toxic conditioning via radiation or chemotherapy currently used for hematopoietic stem cell (HCT) transplantation.

“The possibility of translating these findings into humans is very exciting,” said senior author Seung K. Kim, MD, PhD, a professor of developmental biology at Stanford. “The key steps in our study—which result in animals with a hybrid immune system containing cells from both the donor and the recipient—are already being used in the clinic for other conditions. We believe this approach will be transformative for people with type 1 diabetes or other autoimmune diseases, as well as for those who need solid organ transplants.”

Type 1 diabetes is an autoimmune disease that attacks pancreatic islet cells. While islet transplantation can restore insulin production, it typically requires immunosuppressive drugs that carry risks including infection, malignancy, and organ damage. The Stanford team’s approach reduced these negative effects by inducing immune tolerance through mixed hematopoietic chimerism, a state in which donor and recipient immune cells coexist.

“Mixed hematopoietic chimerism after hematopoietic cell transplantation (HCT) can modulate the immune system and induce tolerance to allogeneic tissues,” the researchers wrote. “However, bone marrow conditioning-related toxicities preclude wider adoption of HCT for transplant allotolerance.”

The current findings by the Stanford team builds on a series of research initiatives beginning with work published in 2022, in which the researchers showed they could cure toxin-induced diabetes in mouse models using antibody-based immune conditioning combined with moderate radiation (200–300 cGy), followed by transplantation of donor-matched blood stem cells and islets. This study served as a proof of concept but used radiation at levels that are potentially toxic.

A November study published in JCI, along with the new research addressed two significant challenges for developing an effective transplantation protocol: autoimmune diabetes, in which the immune system targets islet cells, and the need to reduce conditioning toxicity. In the November study, the researchers added an immune-modulating drug used in autoimmune disease to their regimen. This change enabled the formation of a hybrid immune system that both accepted donor islets and prevented autoimmune attack. All treated mice were protected from developing diabetes, and those with already possessing the disease were cured.

To further reduce toxicity, the April study added additional agents, baricitinib, venetoclax, and an αCD47 antibody, to go with αCD117 antibody and transient T cell depletion. These agents were selected because they target distinct biological pathways involved in immune regulation and bone marrow niche clearance. Baricitinib, a JAK1/2 inhibitor, reduces inflammatory signaling and supports donor cell engraftment. Venetoclax promotes apoptosis of specific immune cells, and αCD47 disrupts a signaling pathway that normally protects cells from clearance, which helped in the removal of host stem cells to make space for donor cells.

“We systematically tested baricitinib (JAK1/2 inhibitor), venetoclax (Bcl2 inhibitor), and αCD47 antibody, agents in current clinical use, and quantified hematopoietic chimerism after HCT,” the researchers wrote. “Combined with αCD117 antibody, transient T cell depletion, and just 10 centigray (cGy) total body irradiation (TBI), these agents enabled durable mixed chimerism and matching allo-islet tolerance, to cure diabetes without evidence of [graft-versus-host disease].”

This new combination allowed researchers to reduce radiation exposure to 10 cGy, a fraction of the levels used in conventional bone marrow transplantation. Mice treated using this regimen showed stable engraftment of donor cells, maintained fertility, and experienced no graft-versus-host disease. They also remained insulin-independent for the duration of the study.

The findings provide a potential pathway toward clinical adoption, which could be speedier than usual since many of the agents used for this approach are already approved or under evaluation in humans.

Work at Stanford will continue in this area and will focus on testing the reduced-intensity regimen in autoimmune diabetes models, refining conditioning strategies, and exploring alternative sources of islet cells, including those derived from stem cells.

If successfully, translated to the clinic, this treatment regimen could reduce or eliminate the need for lifelong insulin therapy, while expanding the use of transplantation-based therapies across a wider set of patients.

The post New Low-Toxicity Transplant Method Reverses Type 1 Diabetes in Mice appeared first on Inside Precision Medicine.

Posted on

Effect of a WeChat Intervention Based on the Common-Sense Model on Breast Cancer–Related Lymphedema Preventive Behaviors: Quasi-Experimental Study

Background: Breast cancer–related lymphedema is the most prevalent postoperative complication among breast cancer survivors. Although mobile health tools are increasingly used for patient education, evidence supporting their efficacy in lymphedema prevention remains limited. Objective: This study aimed to evaluate the effectiveness of a WeChat-based intervention grounded in the common-sense model (CSM) in improving preventive behaviors, modifying illness perceptions, and reducing lymphedema incidence among breast cancer survivors and to validate the targets of the intervention. Methods: This study used a quasi-experimental design. Participants (N=192) were recruited from the breast cancer department of a cancer hospital in Guangzhou, China. The control group (n=98) received routine care. The intervention group (n=94) participated in a 3-month CSM-guided WeChat mini-program (“Nantian e-Care”) delivering tailored educational articles, exercise tutorials, arm circumference monitoring, and real-time nurse consultations. Outcomes, including preventive behaviors, illness perceptions, and lymphedema incidence, were assessed 1, 3, and 6 months post surgery. Generalized estimating equations were used for the analysis. Results: The intervention group exhibited significant improvements in lifestyle adjustments (Wald =6.9, =.03) and physical exercise adherence (Wald =6.9, =.03) compared with the control group. Illness perception, including identity (Wald =8.1, =.04), timeline cyclical (Wald =8.5, =.04), personal control (Wald =9.3, =.03), illness coherence (Wald =29.8, <.001), and behavioral (Wald =19.5, <.001) and physical factors (Wald =24.1, <.001) were markedly enhanced. Mechanistically, skin care improvements were driven by intervention effects, personal control, illness coherence, and behavioral attribution. Lifestyle changes were correlated with intervention and illness coherence. Adherence to physical exercise was not statistically significantly affected by the intervention, although a trend was observed. Critically, the intervention group demonstrated a lower incidence of lymphedema at 6 months (7.50% vs 16.48%, =3.9, =.048). Conclusions: The CSM-guided WeChat intervention effectively promoted preventive behaviors, optimized illness perceptions, and reduced lymphedema risk. These findings underscore the value of integrating theory-driven mobile health tools into postoperative care and highlight scalable strategies for chronic disease management in resource-limited settings. Trial Registration: Chinese Clinical Trial Registry ChiCTR2100048798; https://www.chictr.org.cn/showprojEN.html?proj=130038 International Registered Report Identifier (IRRID): RR2-10.1007/s00520-024-09078-x
Posted on

Optimizing Navigation and Text Messaging Interventions to Promote Participation in a Food Is Medicine Program Among People Participating in Cardiac Rehabilitation: Human-Centered Design Study

Background: Food Is Medicine (FIM) programs integrate interventions such as medically tailored meals or produce prescriptions into clinical care. However, there is limited evidence on how to design these programs to be responsive to the lived experiences of participants to optimize initiation, engagement, and long-term retention. Objective: The objective of the study was to develop interventions to promote initiation, engagement, and retention in FIM programs that are responsive to the lived experiences of participants. Methods: We used a human-centered design approach to engage current and former cardiac rehabilitation participants in the development of interventions to promote participation and engagement in a FIM program. We recruited participants through invitations sent via electronic health record messages. We interviewed participants about their experiences, preferences, and unmet needs related to healthy eating and program design. Additionally, we elicited participant feedback on draft versions of patient navigator scripts and text messages promoting healthy eating habits. Results: A total of six participants identified themes across Theory of Planned Behavior constructs and emergent themes, including the cost of healthy food, cultural appropriateness, clear and timely communication, transportation, local food access, scheduling flexibility, the ability to provide feedback to the program, and personalized support for navigating food resources. Participants described financial strain as a key barrier to healthy eating and noted that social influence often shaped eating behaviors. Feedback on navigator scripts led to revisions clarifying program logistics, addressing barriers such as language and cultural dietary restrictions, and tailoring positive endorsements to individual health goals. Based on participant feedback, text messages were made more concise, reframed positively (eg, humor and gratitude), and encouraged to be warmer, with respectful language that is easy to understand, while avoiding stigmatizing or overly clinical phrasing. Participants also suggested that messages should reflect empathy and offer actionable information to increase trust and engagement with the program. Trust in the health care system and a sense of dignity in receiving food support emerged as critical themes influencing overall satisfaction and retention. Participants emphasized that endorsement from their health care team and cardiologist was important for building trust in the program. Communication between health care navigators and FIM navigators could help reduce the burden placed on patients to navigate food resources. Conclusions: Using a human-centered design approach, we gained insights about participant-identified needs for navigation scripts and text messages that are culturally sensitive and personalized to promote optimal participation in a FIM program.
<img src="https://jmir-production.s3.us-east-2.amazonaws.com/thumbs/106004753afe04c792ee781fb8a4c2ba" />