So, what technology platforms can collaborative team members use to connect with each other without having to leave their desks? High definition video and conferencing apps are an obvious choice.
Much of human communication takes place on the non-verbal level in the form of body language, facial expressions and gestures. These types of non-verbal cues can easily be misunderstood or missed altogether in written communications and so face-to-face interaction is still an essential aspect of successful collaboration.
High definition audio software, as well as group chatting and messaging apps, are also essential items in the collaborative communication toolkit. These enable collaborating staff to quickly and easily communicate and share updates. In addition, social intranets are great enablers of collaborative communications.
With staff biographies, follow-me functionality and individual or project activity walls, team members will be able to keep up to date with the latest developments. As well as being very effective, these tools are also inexpensive, which will be a big plus when making the case to the Director of Finance! File sharing tools are another essential item in the collaboration toolbox.
Being able to quickly and easily share documents and files is essential to the smooth functioning of collaborative communication. And there are a number of software platforms out there that you can choose from. Dropbox is perhaps the most well known. It may be that you are also using OneDrive or other file sharing apps like WeTransfer.
And while these apps are certainly convenient and will do the job for you, they also have a number of disadvantages. There are, for instance, question marks over security. SSL encryption and folder locking features are perhaps not as robust as they could be. Perhaps the most important disadvantage though is an inability to edit and collaborate on documents in real-time.
In comparison, cloud intranets offer the same ease of use and convenience but with the added advantages of price certainty, security assurance and unlimited storage. You see, cloud intranet providers retain responsibility for the security of all company information and data. They use the very latest encryption techniques and security features to safeguard important or commercially sensitive documents from cybercrime and hacking.
Along with automatic backups and updates, a cloud intranet also comes with unlimited user numbers and storage all for just a modest monthly fee with no hidden extras. Keeping on top of where things are at is not easy. Cloud technology, especially in the shape of an intranet, will do away with these challenges resulting in more streamlined working processes.
As well as making it easier to share files and documents both internally and externally, team members can simultaneously work on the same document in real-time via an intranet-embedded G Suite or Office app. Collaborative team members will no longer spend valuable time navigating complicated email threads for the most up to date version of the document. Nor will they need to make sense of complex track changes and editing comments.
Instead, they can have absolute confidence that the intranet has the very latest version of the document. Embedding the document or spreadsheet directly in the intranet also means that essential contextual or background information can be displayed alongside. So, if you want easy, convenient and fast file and document sharing that also allows for real-time editing and collaboration, then a cloud intranet may well be the platform for you. In addition, the cloud intranet offers the opportunity for team members to brainstorm ideas, easily share insights and data either through a collaborative workspace or a project blog.
Access can be restricted to only the relevant team members through a series of intranet site permissions. Workspaces or blogs can transcend time and location and enable employees to connect, share and collaborate effectively. These are another very effective collaboration tool. Team members can share their entire desktop screen, individual window or specific application with one or more colleagues as well as streaming videos or presentations.
Shared-screen apps are a great way for collaborative team members to show their colleagues exactly what they are talking about and help to keep everyone on the same page as well as ensuring meetings stay on track.
Project management systems or task management software will be another important weapon in the collaborative arsenal. And many of these tools require extensive staff training before employees can get the best out of them. Indeed, there may well be a sharp intake of breath when you see the price tag at the end of it all! A cloud intranet, on the other hand, can easily be used to develop a streamlined project management process. From project initiation and planning through to execution, launch and control, an intranet will support operations with online processes offering central oversight.
The ease of communication made possible by the intranet is one obvious advantage. And so, if your business is serious about promoting collaboration, then the starting point must be communications. The broad pattern of the induced SST anomalies is also not unlike the one seen in the observations for the positive phase of the NAO 16 , Anomalies are with respect to the pre-industrial PI experiment. The dashed lines are linear regression curves with the correlation coefficient r and the regression slopes k also shown.
In principle, the subpolar cooling can be explained by two effects induced by intensified surface westerly winds Fig. First, the wind intensification cools the ocean surface by enhancing turbulent heat fluxes to the atmosphere. However, as we argued earlier for the observations, the impact of surface wind intensification is not evident in actual turbulent heat flux changes, which are positive i.
Second, surface westerly wind anomalies drive an anomalous southward Ekman transport of cold water, resulting in a balance between anomalous ocean heat divergence and the reduction of surface heat fluxes.
A year running mean is applied. The general warming in the subtropical Atlantic is consistent with the easterly winds-induced northward Ekman transport of warm water. As a result of all these aforementioned changes in the North Atlantic, the NAWH index experiences a rapid reduction within a few years by about 0. In short, the Indian Ocean warming suppresses the Atlantic rainfall by modifying the Walker circulation along the entire equatorial belt, and the resultant positive Atlantic salinity anomalies act to strengthen the AMOC after being advected to the subpolar North Atlantic by ocean mean circulation.
At equilibrium, the stronger westerly winds over the subpolar region also help maintain a stronger AMOC by enhancing surface turbulent heat loss. The AMOC plays an important role in the global interhemispheric energy transport and its net effect is to carry ocean heat northward towards the subpolar North Atlantic.
The initial cooling south of Greenland gradually vanishes and is later replaced by a basinwide warming Fig. The slow response, as indicated by the difference between Years — and Years 1—40, features a pronounced warming to the south of Greenland caused mainly by the AMOC strengthening Fig. It is interesting to note that the large-scale wind pattern at the final stage does not change much from the fast response Fig.
For the observations, specifically for the tight correlation shown in Fig. Further, we conduct a series of sensitivity experiments to confirm the robustness of the results Methods , and here we discuss several of them.
As part of the fast response, the surface wind field shows a negative NAO-like pattern with anomalous easterly winds over the subpolar North Atlantic and westerly winds on its southern flank Fig. In a direct response to the wind changes, the subpolar North Atlantic warms up while subtropical North Atlantic cools.
For example, the anomalous subpolar westerly winds and subtropical easterly winds are both evident in the fast response Fig. In the fast response, the North Atlantic has a suppressed warming in the subpolar region Fig. In Fig. For the slow response associated with the remotely induced AMOC changes, the correlation is about 0.
Colored triangles indicate the observed changes during — long-term trends multiplied by 66 years for different datasets; the black triangle represents their averages. The dashed lines are linear regression curves with the correlation coefficients r and the regression slopes k also shown.
The net impact on the NAWH is a combination of both, and its sign and magnitude depend on the relative magnitude of the fast and slow responses. Based on our model simulations, the sensitivities of the NAWH index to the relative TIO warming for the fast and slow responses are overall comparable, which makes their net long-term impact smaller due to strong compensation. It is noteworthy however that the NAWH properties strength, location, etc.
Given the relatively long timescale of the AMOC response to TIO warming and no observational evidence for AMOC strengthening, the fast response appears to be more relevant to explaining the trends of the past half-a-century or so. Thus, the general similarities in the spatial patterns and our rough estimates above suggest that atmospheric changes related to the Indian Ocean warming can at least in part account for the observed NAWH. However, our results also indicate that on multidecadal and longer timescales atmospheric and oceanic effects of the TIO warming on the NAWH would nearly compensate each other, unless other factors act to weaken the AMOC.
In fact, in future climate projections the NAWH appears to be maintained predominantly by the robust AMOC decline 3 , 6 , 9 caused by increased ocean stratification in the high-latitude North Atlantic. For the late 20th century and the last two decades tentative evidence suggests that the AMOC might be already weakening 4 , 5 , 33 , 34 , but the data is either too short or based on proxy reconstructions.
However, our results and the results of ref. However, others argue that these lead-lag correlations do not necessarily shed light on causality because on such timescales SST is already in a quasi-equilibrium with the atmosphere Moreover, our results suggest that the atmospheric and oceanic mechanisms of the NAWH are not necessarily exclusive of each other, and both could operate at the same time.
Thus, the relative contribution of atmospheric changes versus AMOC changes to the observed NAWH needs to be investigated and quantified in future studies.
This index is designed to capture regional SST deviation in the subpolar North Atlantic from the surrounding North Atlantic; such regional deviation is more relevant to the concept of the warming hole than the subpolar SST change. Ocean and atmosphere are fully coupled globally except for the region where we perturb ocean surface temperature. We firstly conduct idealized simulations to demonstrate the coupled system response to the tropical Indian Ocean warming.
All the simulations are integrated for years unless specifically mentioned. PI refers to the control, pre-industrial simulation. More details on the experimental set-up can be found in ref. Subsequently, the AMOC strength changes little in the first three decades, but then starts to increase as the salinity anomalies are transported northward. Hansen, J. Global surface temperature change. Drijfhout, S. ADS Google Scholar.
Menary, M. Google Scholar. Rahmstorf, S. Exceptional twentieth-century slowdown in Atlantic Ocean overturning circulation. Change 5 , Arctic sea-ice decline weakens the Atlantic meridional overturning circulation. Change 7 , — Gervais, M. Hu, S. Global pattern formation of net ocean surface heat flux response to greenhouse warming.
Liu, W. The role of the Atlantic meridional overturning circulation in a warming climate. Chemke, R. Identifying a human signal in the North Atlantic warming hole. Clement, A. Testing the role of the ocean in historical simulations of Atlantic multidecadal variability and the North Atlantic warming hole. Josey, S. Extreme variability in Irminger Sea winter heat loss revealed by ocean observatories initiative mooring and the ERA5 reanalysis. Lozier, S. Hoerling, M. Tropical origins for recent North Atlantic climate change.
Science , 90—92 Hurrell, J. Twentieth century North Atlantic climate change. Part I: assessing determinism. Twentieth century North Atlantic climate change: II.
Marshall, J. A study of the interaction of the North Atlantic Oscillation with ocean circulation. Alory, G. Observed temperature trends in the Indian Ocean over — and associated mechanisms. Du, Y. Most of the signatures were present across all age groups, and represent the following biological processes in order of abundance in neuroblastoma: 1 signature 18, which has been causally associated with reactive oxygen species ROS -induced mutagenesis in cultured iPS cells and mouse models 42 , 43 ; 2 signature 3, putatively caused by homologous recombination deficiency; 3 and 4 clock-like signatures 5 and 1, indicative of unclear causes and 5-methylcytosine deamination, respectively; 5 signature 31, which is cisplatin-induced and was found exclusively at relapse in 5 of 23 relapses 41 , 44 ; and 6 signature 40, of unknown etiology and found in a single sample Fig.
Sample numbers are: group A, 39; B, ; and C, Shared mutations present in both diagnosis and relapse, which are thus early mutations are indicated at left, followed by relapse-specific variants detected only at relapse on the right for each patient.
Y -axis represents -log 10 P values; x -axis represents the log expression fold-change mean expression difference , with genes increased in signature positive samples in the positive direction red and genes increased in signature negative samples in the negative direction blue. Each point represents one gene, and genes with the lowest 1.
Interquartile range, median, and whiskers are as in d. Overlaid are, from left to right: 1 ssGSEA Z -scores for the neural gene set Supplementary Table 2 with red indicating more neural expression, 2 the 17q gain status with purple indicating no gain, and 3 proportion of mutations caused by signature 18 in each sample Z -scores.
Dotted outline is to enable comparison of a neural-enriched group between each plot. Signature 18 has predominantly been detected in neuroblastoma but is rare in most other cancers 41 and is induced by ROS in some model systems 42 , We found that signature 18 may be both an early event in neuroblastoma, causing truncal mutations, and an on-going mutational event causing relapse-specific mutations, based on analysis of five patients with matched diagnosis and relapse samples Fig.
Four of the five patients had signature 18 in both shared early and relapse-specific late mutations, while the fifth lacked the signature at both diagnosis and relapse Fig. This suggests an intrinsic and stable propensity of specific neuroblastoma tumors to either possess or lack the signature 18 mutational process. The prevalence of signature 18 varied considerably, ranging from 0 to SNVs per tumor Supplementary Fig.
To identify biological processes that may contribute to this variability, we compared gene expression profiles between samples with detectable signature 18 vs. This analysis included 88 diagnosis samples that had both WGS and RNA-Seq, 28 of which were signature negative and 60 of which were signature positive.
Differential expression was analyzed using Limma to generate Benjamini-Hochberg-adjusted P values for each gene Fig. Genes involved in neural function, including PIRT affecting peripheral nerve function 48 , TMEFF2 which promotes neuron survival 49 , DST a neural cytoskeletal gene 50 , and GABBR1 a neurotransmitter receptor 51 , were among the most significantly increased in expression in signature negative samples.
Indeed, Indeed, 17q gain itself was significantly associated with increased signature 18 proportion Fig. Further, the neural-enriched samples, with neural expression scores determined using ssGSEA 55 of 26 neural genes statistically increased in signature negative samples Fig. These data suggest that the neural-enriched samples lacking signature 18 may represent a distinct subgroup; indeed, a neuronally differentiated transcriptional subgroup was previously identified in neuroblastoma We also tested whether other genetic alterations, in addition to 17q gain, were associated with signature 18 Supplementary Fig.
MYCN -altered samples had significantly more signature 18 Fig. MYCN and 17q gains were also statistically enriched in signature positive samples when including only stage 4 samples Supplementary Fig.
Further, samples with both 17q gain and MYCN alterations had even higher signature 18 than those with 17q gain alone, suggesting additive interaction Supplementary Fig.
Finally, MYCN was not the primary driver of increased mitochondrial gene expression in signature positive samples, since mitochondrial genes remained statistically increased when differential gene expression analysis included MYCN alteration status as a covariate, to remove potential confounding effects of MYCN Supplementary Fig.
This is consistent with 17q gains potentially promoting the mitochondrial gene expression, given that several of the mitochondrial genes are found on 17q. Indeed, when both 17q gains and MYCN status were included as covariates, signature 18 alone was not significantly associated with expression of any gene, and mitochondrial gene expression was associated with 17q gains Supplementary Fig.
This indicates that 17q gains were the primary factor associating signature 18 with mitochondrial gene expression. To test whether similar signature 18 correlations could be observed in other cancer types, we analyzed mutational signatures in pediatric cancers with WGS spanning 39 cancer types including all major pediatric cancers, and adult cancers from TCGA 58 with WGS spanning 24 cancer types.
As only three signature positive liver tumors were available, and the colon epithelium is subjected to continuous environmental exposures which may confound the search for endogenous signature causes, we focused analysis on rhabdomyosarcoma which had 13 signature positive samples out of 31 samples total, sequenced by both WGS and RNA-Seq. We performed differential gene expression analysis with Limma, comparing signature positive vs. The specific nuclear-encoded mitochondrial genes in rhabdomyosarcoma were different from those enriched in neuroblastoma signature positive samples, except for COX8A Fig.
Together, these results suggest that certain copy gains e. However, experimental data are ultimately needed to test this hypothesis. We next analyzed which driver SNVs were likely induced by signature We quantified the probability that driver mutations in these genes along with BRAF , TP53 , and MYCN were induced by each mutational signature using an approach described previously 62 which we have applied in other pediatric cancers 63 Fig.
Among the WGS samples with signature reconstruction cosine similarity of 0. This indicates that signature 18 is likely a driver of disease progression in neuroblastoma, in contrast with passenger mutational signatures, such as the kataegis-associated APOBEC signature in osteosarcoma which causes no known driver SNVs in that cancer type Each row represents one mutation in a specific patient. Trinucleotide context of each mutation is indicated at right. Genes are color-coded as indicated in legend at bottom-right.
We previously reported age-associated ATRX alterations in a cohort of neuroblastoma patients These findings suggest that the sympathetic nervous system, the tissue from which neuroblastoma arises, is susceptible to different oncogenic insults at different times during development, which could be explored in future investigations using animal models.
This may be due to 1 age-specific oncogene dependencies in the tissue of origin, or 2 different proliferation rates conferred by specific alterations, leading to clinically detectable disease at earlier or later ages. ALK N-terminal variants, present in 5 of 2. While attempts to inhibit ALK in neuroblastoma have been disappointing 66 , these efforts have focused on patients with kinase domain point mutations. Since the kinase domain is preserved in the ALK N-terminal variants, functional and pharmacological studies are merited to determine whether ALK kinase inhibitors are effective against these alterations.
Our findings indicate that the signature 18 mutational process is the most common cause of driver SNVs in neuroblastoma. This suggests that this mutagenic process, which is caused by ROS in other settings 42 , 43 though not proven in neuroblastoma , may promote evolution and heterogeneity, as many driver SNVs, such as ALK mutations, are later events in neuroblastoma 11 , Mitochondrial ribosomal and electron transport chain gene expression was associated with higher signature 18 in neuroblastoma—a finding that we were able to replicate in rhabdomyosarcoma Fig.
Indeed, mitochondrial electron transport chain components I and III are known to generate ROS 61 , a reported cause of signature 18 in certain experimental models 42 , 43 , and increased mitochondrial ribosome proteins may promote increased translation of complex I, some of whose subunits are translated by mitochondrial ribosomes We speculate that the poor outcomes in neuroblastomas with 17q gain may be partly due to their increased mutational burden Fig. This in turn may cause ROS production and potentially signature induced mutagenesis, which fuels point mutations and clonal heterogeneity promoting drug resistance.
Experimental models are ultimately needed to test these hypotheses. Indeed, mitochondrial metabolism has also been suggested to play a role in neuroblastoma pathogenesis by others Together, our findings suggest that there may be therapeutic vulnerabilities in neuroblastomas with 17q gain, potentially through targeting altered mitochondrial function.
Jude Clinical Genomics samples were obtained from St. Jude under IRB approval from that institution. This study complies with the Declaration of Helsinki and all other relevant ethical regulations. All statistical analyses were performed using two-sided tests if applicable. Specific tests used are described in the main text or figure legends. When comparing distributions we used the non-parametric Wilcoxon rank-sum test due to frequently skewed distributions.
Each sample was sequenced one time and no technical replicates were performed. Jude Clinical Genomics programs.
CNV data were manually inspected by plotting copy changes and LOH across the genome for each sample, and if necessary, CNV data were manually centered with presumed diploid regions at copy level 2. Presumed diploid regions were considered to be those at the lowest non-LOH-region copy level, and the resulting presumed haploid and triploid regions obligately having LOH. A median level of 2. In addition, the median copy value for the entire opposing control arm had to be at least 0.
However, for 11q If both arms of the chromosome met this criterion, the whole chromosome was considered gained. WES-based arm-level copy alterations. WES-based arm-level copy calls with CnvKit had CNVs in each tumor were integrated into the structural variant analysis by matching breakpoints within a 5-kb window to rescue rearrangements with CNV support by manual curation. Somatic SNVs and indels were called using Bambino version 1.
They were likely contamination from another neuroblastoma sample with MYCN amplification run on the same sequencing lane, as the large amount of MYCN DNA in amplified samples may increase the potential for artefactual spill-over errors between samples run on the same lane. Novel junctions were analyzed by RNApeg version 1 Genes with at least 1 count per million CPM in at least two samples were included, and RNA-Seq count data were transformed using voom 74 in R which log-transforms data and performs statistical analysis preparatory to linear modeling , followed by the Limma 75 functions lmFit for linear modeling , eBayes for Bayesian differential expression analysis , and topTable to report top differentially expressed genes ; default parameters were used for each of these functions, including Benjamini—Hochberg correction to obtain adjusted P values.
A variant of this analysis was also performed with the inclusion of somatic MYCN alteration status amplifications or point mutations were both considered positive as a covariate to subtract potential effects of MYCN in the Limma differential expression analysis Supplementary Fig. Rhabdomyosarcoma data are available through the St. Rhabdomyosarcoma differential gene expression was performed using similar steps to the neuroblastoma analysis, except that 1 two batches were included as covariates since some samples utilized poly-A-based unstranded library preparation and some used total RNA stranded library preparation, 2 the robust option was used with lmFit, due to the smaller sample size, to be robust against outliers, and 3 genes were more stringently filtered due to the smaller sample size, including genes with at least 10 CPM in at least 10 samples including over 11, genes.
We tested mutual exclusivity and co-occurrence of recurrent mutations, using diagnosis WGS samples, including only recurrent mutations mutated in five or more samples in that sample set. We performed this analysis after removing non-independent interactions.
SNVs and indels were always classified as independent from all other events, while translocations and copy number alterations were considered non-independent if a a structural variant joined them directly, as in the case of t 11;17 variants that directly joined to the end of an 11q To determine whether copy number gains were an early or late event in neuroblastoma evolution, we used mutations in regions with 3 copies as this was the most frequent level of copy gain observed, and also is the easiest to model as it has fewer possible allelic configuration than higher copy gains in out of the diagnosis samples.
The expected VAF for mutations on 2 of 3 copies 0. Mutagenesis after the copy gain will occur on 1 of 3 copies, such that the percentage of mutations on 1 of 3 alleles will increase over time.
In Supplementary Fig. We applied Cis-X 77 version 1. We then applied SigProfiler 44 version 2. The optimal number of extracted signatures was 6, which had a signature stability above 0. The sixth, T, was highly similar cosine similarity of 0. We then used SigProfilerSingleSample version 1.
Signatures 12, 43, and 46, which were initially detected by de novo signature extraction, were not detected by SigProfilerSingleSample in any sample, leaving 7 signatures including T detected in our final analysis. To identify variants likely caused by specific mutational signatures, we used an approach 62 which we have implemented previously 63 , as follows.
The probability that an SNV was caused by a specific signature was calculated as follows. The probability that a mutation of interest m at trinucleotide context c was caused by a specific signature i is calculated as:. The numerator represents the number of mutations caused by a specific signature i at the mutation context of interest, while the denominator represents the total number of mutations caused by all signatures detected in the sample at the mutation context of interest.
Example calculations for a specific mutation are shown in Supplementary Fig. Jude Clinical Genomics programs available on St. Jude Cloud. The pediatric cancer data use the GRCh38 reference genome. Signature scores were obtained using SigProfilerSingleSample. Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Source data are provided with this paper. Any remaining data are available within the Article, Supplementary files, or are available from the authors upon request. Maris, J. Lancet , — Cohn, S. Simon, T. New definition of low-risk neuroblastoma using stage, age, and 1p and MYCN status. PubMed Google Scholar. Smith, V. High-risk neuroblastoma treatment review. Children 5 , Baker, D. Outcome after reduced chemotherapy for intermediate-risk neuroblastoma.
New Engl. Schleiermacher, G. Cancer 97 , — Janoueix-Lerosey, I. Overall genomic pattern is a predictor of outcome in neuroblastoma. PubMed Article Google Scholar.
Lopez, G. Somatic structural variation targets neurodevelopmental genes and identifies SHANK2 as a tumor suppressor in neuroblastoma. Genome Res. Valentijn, L. TERT rearrangements are frequent in neuroblastoma and identify aggressive tumors. Cheung, N. Association of age at diagnosis and genetic mutations in patients with neuroblastoma. JAMA , Ma, X. Pan-cancer genome and transcriptome analyses of 1, paediatric leukaemias and solid tumours.
Nature , — Heaphy, C. Science , — Identification of ALK as a major familial neuroblastoma predisposition gene. George, R. Activating mutations in ALK provide a therapeutic target in neuroblastoma. Okubo, J. Oncogene 31 , — Cazes, A. Characterization of rearrangements involving the ALK gene reveals a novel truncated form associated with tumor aggressiveness in neuroblastoma.
0コメント