《自然》（20190425出版）一周論文導讀

翻譯 | 宗華

Nature, 25 April 2019, Volume 568 Issue 7753

《自然》2019年4月25日，第7753期568卷

人工智慧Artificial Intelligence

Speech synthesis from neural decoding of spoken sentences

來自口語句子神經解碼的語言合成

作者：Gopala K. Anumanchipalli, Josh Chartier & EdwardF. Chang

鏈接：

https://www.nature.com/articles/s41586-019-1119-1

摘要：

將神經活動轉化為語言的技術，將對那些由於神經功能受損而無法交流的人產生革命性的影響。在這裡，我們設計了一個神經解碼器。其能利用人類皮層活動中編碼的運動學和聲音表徵來合成可聽語音。

遞歸神經網路首先將直接記錄的皮層活動解碼為發音運動的表徵，然後將這些表徵轉化為語音聲學。在封閉辭彙測試中，聽者可以很容易地識別和轉錄由大腦皮層活動合成的語音。

中間發音動力學增強了性能，即便數據有限。解碼的發音表徵在不同的說話者之間高度保守，使得解碼器的一個組件可在不同的參與者之間傳輸。

此外，當參與者默寫句子時，解碼器可以合成語音。

Abstract

Technology that translates neural activity into speech would be transformative for people who are unable to communicate as a result of neurological impairments. Here we designed a neural decoder that explicitly leverages kinematic and sound representations encoded in human cortical activity to synthesize audible speech. Recurrent neural networks first decoded directly recorded cortical activity into representations of articulatory movement, and then transformed these representations into speech acoustics. In closed vocabulary tests, listeners could readily identify and transcribe speech synthesized from cortical activity. Intermediate articulatory dynamics enhanced performance even with limited data. Decoded articulatory representations were highly conserved across speakers, enabling a component of the decoder to be transferrable across participants. Furthermore, the decoder could synthesize speech when a participant silently mimed sentences.

Predicting disruptive instabilities in controlled fusion plasmas through deep learning

通過深度學習預測受控聚變等離子體的破裂不穩定性

作者：Julian Kates-Harbeck, Alexey Svyatkovskiy & William Tang

鏈接：

https://www.nature.com/articles/s41586-019-1116-4

摘要：

由磁約束托卡馬克反應堆提供的核聚變能源有望成為可持續的清潔能源。避免這些反應堆發生大規模等離子體不穩定性，即所謂的「破裂」，是最緊迫的挑戰之一，因為破裂可能導致電力生產中斷，並損壞關鍵部件。

這裡，我們提出了一種基於深度學習的破裂預測方法。我們的方法在很大程度上擴展了以前策略的能力，例如基於第一性原理和經典機器學習的方法。

我們的方法利用高維訓練數據提高預測性能，同時最大限度地利用超級計算資源提高準確性和速度。

Abstract

Nuclear fusion power delivered bymagnetic-confinement to kamak reactors holds the promise of sustainable and clean energy. The avoidance of large-scale plasma instabilities called disruptions within these reactors is one of the most pressing challenges, because disruptions can halt power production and damage key components. Here we present a method based on deep learning for forecasting disruptions. Our method extends considerably the capabilities of previous strategies such as first-principles-based and classical machine-learning approaches. Our approach takes advantage of high-dimensional training data to boost predictive performance while also engaging supercomputing resources at the largest scale to improve accuracy and speed.

天文學Astronomy

Martian dust storm impact on atmosphericH₂Oand D/H observed by ExoMars Trace Gas Orbiter

外火星跟蹤氣體軌道器觀測到的火星塵暴對大氣H₂O和D/H 的影響

作者：Ann Carine Vandaele, Oleg Korablev, Frank Daerden, et al

鏈接：

https://www.nature.com/articles/s41586-019-1097-3

摘要：

全火星的沙塵暴雖然少見，但可以影響火星大氣幾個月。這裡，我們提出了對全火星沙塵暴開始時的塵埃、水和半重水（HDO）進行的同步、高解析度測量。

測量結果由外火星跟蹤氣體軌道器上的NOMAD和ACS儀器獲得。我們報告了從行星邊界層到80千米高空的HDO/H₂O比（D/H）的垂直分布。

研究結果表明，在沙塵暴開始前，HDO丰度在40千米以上的高度下降至可檢測水平以下。HDO的減少與水冰雲的存在是一致的。沙塵暴期間，在40~80千米高度觀察到H₂O和HDO的丰度增加。

我們認為，這些丰度的增加可能是沙塵暴期間溫度升高的結果，導致更強的大氣環流和防止冰雲形成。這可能通過重力下降和冰晶的隨後升華將水蒸氣限制在較低的高度。

Abstract

Global dust storms on Mars are rare but can affect the Martian atmosphere for several months. Here we present concurrent, high-resolution measurements of dust, water and semiheavy water (HDO) at the onset of a global dust storm, obtained by the NOMAD and ACS instruments on board the Exo Mars Trace Gas Orbiter. We report the vertical distribution of the HDO/H₂O ratio (D/H) from the planetary boundary layer up to an altitude of 80 kilometres. Our findings suggest that before the onset of the dust storm, HDO abundances were reduced to levels below detectability at altitudes above 40 kilometres. This decrease in HDO coincided with the presence of water-iceclouds. During the storm, an increase in the abundance of H₂O and HDO wasobserved at altitudes between 40 and 80 kilometres. We propose that these increased abundances may be the result of warmer temperatures during the dust storm causing stronger atmospheric circulation and preventing ice cloud formation, which may confine water vapour to lower altitudes through gravitational fall and subsequent sublimation of ice crystals.

No detection of methane on Mars from early ExoMars Trace Gas Orbiter observations

外火星氣體軌道器早期觀測未發現火星上有甲烷

作者：Oleg Korablev, Ann Carine Vandaele, Franck Montmessin, et al

鏈接：

https://www.nature.com/articles/s41586-019-1096-4

摘要：

火星上甲烷的發現被解釋為表明地球化學或生物活動可能在今天的火星上繼續存在。

2018年4~8月，歐空局—俄羅斯太空署外火星跟蹤氣體軌道器上搭載的ACS和NOMAD儀器，對火星大氣進行了高度敏感的測量，以試圖探測甲烷。

我們未在兩個半球的一系列緯度範圍內探測到甲烷，同時得到的甲烷體積上限約為0.05ppm，是之前報道的陽性檢測結果的1/100~1/10。

我們認為，實現在蓋爾隕石坑中發現的背景甲烷濃度與目前的發現之間的一致，需要一個未知的過程。該過程能在甲烷在整個火星擴散前迅速從低層大氣中將其移除或隔離。

Abstract

The detection of methane on Mars has been interpreted as indicating that geochemical or biotic activities could persiston Mars today. Here we report highly sensitive measurements of the atmosphere of Mars in an attempt to detect methane, using the ACS and NOMAD instruments onboard the ESA-Roscosmos ExoMars Trace Gas Orbiter from April to August 2018. We did not detect any methane over a range of latitudes in both hemispheres, obtaining an upper limit for methane of about 0.05 parts per billion by volume, which is 10 to 100 times lower than previously reported positive detections. We suggest that reconciliation between the present findings and the background methane concentrations found in the Gale crater would require an unknown process that can rapidly remove or sequester methane from the lower atmosphere before it spreads globally.

基因組學Genomics

New insights from uncultivated genomes ofthe global human gut microbiome

來自全球人類腸道微生物群的未培養基因組的新見解

作者：Stephen Nayfach, Zhou Jason Shi, Rekha Seshadri, et al

鏈接：

https://www.nature.com/articles/s41586-019-1058-x

摘要：

由於在實驗室條件下培養微生物面臨挑戰，很多人類腸道微生物的基因組序列仍然未知。在這裡，我們通過從3810個排泄物宏基因組中重建60664個原核基因組草圖解決這個問題。

這些基因組為2058個新鑒定的物種水平的操作分類學單位（OTUs）提供了參考點。這意味著比以前已知的已測序腸道細菌的系統發育多樣性增加了50%。

平均而言，新發現的OTUs占每個個體豐富度的33%和物種丰度的28%，並且在農村人口中更加豐富。

一項對臨床腸道微生物研究的薈萃分析指出，很多疾病與新發現的OTUs有關，而這有可能改進預測模型。

最後，我們的分析顯示，未培養的腸道物種已經經歷了基因組的減少，導致某些生物合成途徑的丟失。這可能為今後改進培養策略提供線索。

Abstract

The genome sequences of many species of the human gut microbiome remain unknown, largely owing to challenges in cultivating micro organisms under laboratory conditions. Here we address this problem by reconstructing 60,664 draft prokaryotic genomes from 3,810 faecal metagenomes, from geographically and phenotypically diverse humans. These genomes provide reference points for 2,058 newly identified species-level operational taxonomic units (OTUs), which represents a 50% increase over the previously known phylogenetic diversity of sequenced gut bacteria. On average, the newly identified OTUs comprise 33% of richness and 28% of species abundance perindividual, and are enriched in humans from rural populations. A meta-analysisof clinical gut-microbiome studies pinpointed numerous disease associations for the newly identified OTUs, which have the potential to improve predictive models. Finally, our analysis revealed that uncultured gut species have undergone genome reduction that has resulted in the loss of certain biosynthetic pathways, which may offer clues for improving cultivation strategies in the future.

A new genomic blueprint of the human gutmicrobiota

人類腸道微生物群的新基因組藍圖

作者：Alexandre Almeida, Alex L. Mitchell, Miguel Boland, et al

鏈接：

https://www.nature.com/articles/s41586-019-0965-1

摘要：

人類腸道微生物群的組成與健康和疾病有關，但要解釋它們的生物學作用，需要了解單個微生物物種。

我們通過從11850個人類腸道微生物群中重建92143個元基因組，鑒定了1952種未培養的候選細菌。這些未培養的基因組大大擴展了人類腸道微生物群的已知物種庫，使它們的系統發育多樣性增加了281%。

儘管與參考分離基因組相比，新發現的物種在得以廣泛研究的種群中不那麼普遍，但它們改善了未得到充分研究的非洲和南美樣本的分類，使其多樣性提高了200%以上。

這些候選物種編碼了數百個新發現的生物合成基因簇，並且具有獨特的功能能力。這可能解釋了它們難以捉摸的本質。

Abstract

The composition of the human gut microbiotais linked to health and disease, but knowledge of individual microbial speciesis needed to decipher their biological roles. Here we identify 1,952 uncultured candidate bacterial species by reconstructing 92,143 metagenome-assembled genomes from 11,850 human gut microbiomes. These uncultured genomes substantially expand the known species repertoire of the collective human gutmicrobiota, with a 281% increase in phylogenetic diversity. Although the newly identified species are less prevalent in well-studied populations compared to reference isolate genomes, they improve classification of understudied African and South American samples by more than 200%. These candidate species encode hundreds of newly identified biosynthetic gene clusters and possess adistinctive functional capacity that might explain their elusive nature.

關注我們

微信號：sciencenet-cas(←長按複製) 或長按下方二維碼

喜歡這篇文章嗎？立刻分享出去讓更多人知道吧！