Key Concept: Synthetic Lethality

Synthetic lethality arises where cell death is produced by loss or inhibition of function events in two or more genes - and where cells remain viable when any one of these events occurs in isolation [O'Neil N, et al. 2017, Wappett M, et al. 2016]. Therefore, synthetic lethality can be exploited to kill cancer cells; a classic example is where mutations in BRCA1 or BRCA2, part of the homologous recombination DNA repair pathway, result in dependency on DNA repair by PARP genes, making cells exquisitely sensitive to pharmacological inhibition of PARP1/2 [Lord C & Ashworth A, 2017, Bryant H, et al. 2005, McCabe N, et al. 2006].



Therefore, mutually exclusive loss is a key consequence of synthetic lethal relationships. SynLeGG enables exploration of mutually exclusive, or depleted, gene states in polyomics data across a large number of cell lines and tissues. In addition to having therapeutic potential, these dependencies may reveal gene functions.

Discovery of Gene Expression Clusters with MultiSEp

SynLeGG presents results from the MultiSEp algorithm for RNA-seq data from a comprehensive cancer cell line atlas (CCLE) [Ghandi M et al. 2019]. MultiSEp applies Gaussian mixture modelling with expectation-maximisation to discover gene expression clusters, or modes, where cardinality is determined by Bayesian Information Criterion regularisation [Lubbock A et al. 2013]. This method allows categorisation of cell lines into different clusters based on the expression of a given gene, flexibly capturing between two to five different sub-populations of cell lines (Figures 1-3, below). Each cell line is assigned a probability of belonging to each cluster by MultiSEp, which determines cluster assignment; for example when a cell line is positioned in a region of overlap between two clusters.

Figure 1. Two gene expression clusters identified by MultiSEp. A bimodal distribution is shown, where MultiSEp identified two clusters of cell lines from the CCLE data for EAF2. The first cluster is coloured red; the second cluster is shown in blue. Each of the two clusters could potentially be split further, however MultiSEp found that the bimodal model was top-scoring when compared to the the other models evaluated (from 2 to 5 clusters, scored using Bayesian Information Criterion). Therefore EAF2 gene expression was divided into two clusters.

Figure 2. Three gene expression clusters identified by MultiSEp. MultiSEp identified three clusters (modes) of cell lines from the CCLE data for CDK6. The first cluster is coloured red; the second cluster is shown in green, the third in blue. Each of the three clusters could potentially be split further, however MultiSEp determined that the trimodal model was top-scoring when compared to the the others evaluated (from 2 to 5 clusters, scored using Bayesian Information Criterion). Therefore three clusters were assigned to the CDK6 gene expression data.

Figure 3. Four gene expression clusters identified by MultiSEp. MultiSEp identified four clusters (modes) of cell lines from the CCLE data for FERMT3. The first cluster is coloured red; the second, third and fourth clusters are respectively shown in green, blue and purple. Each of the four clusters could potentially be split further, however MultiSEp determined that the above model was top-scoring when compared to the the others evaluated (from 2 to 5 clusters, scored using Bayesian Information Criterion). Therefore four clusters were assigned to the FERMT3 gene expression data.

The cell line by gene, cluster assigment matrix is available to download here

Integration of Gene Expression Clusters with CRISPR Screen Data

Gene expression clusters enable partitioning of effect scores from CRISPR screen data, and therefore may reveal pairwise gene dependencies. SynLeGG visualises MultiSEp mRNA expression clusters and the CRISPR scores from CERES, including statistical evaluation of dependency relationships (Figure 4). A log2 fold-change and two-tailed T-test p-value are calculated between the CRISPR scores for each consecutive pair of mRNA expression clusters. Only gene pairs with log2 fold-change values less than -0.1 between clusters and p-values less than 0.1 are returned, q-values are also provided [Benjamini & Hochberg 1995]. We strongly recommend taking the q-values as a baseline for determining statistical significance, indeed it is crucially important to correct for multiple hypothesis testing to ensure the statistical validity of results. Users that are following up a hypothesis for a single gene pair that has arisen in analysis of data from outside of SynLeGG might have grounds to refer to the trends described by the uncorrected p-values. SynLeGG also allows the user to explore tissue-specific penetrance (Figure 5).

Figure 4. Visualisation of NMT1 CRISPR scores by NMT2 gene expression clusters for all tissues represented in CCLE. Each dot represents a cell line and the tissue of origin is shown in the key. The NMT1 CRISPR score is shown on the y-axis, NMT2 gene expression clusters are shown on the x-axis. Lower CRISPR scores represent greater depletion with NMT1 loss of function and so correspond to a larger effect upon cell viability/growth. The CRISPR scores are shifted and scaled per cell line so that nonessential genes have a median score of 0 and essential genes have a median score of -1. The NMT1 CRISPR scores are lower in the first NMT2 expression cluster (left), relative to the second expression cluster (right). Therefore, low expression of NMT2 appears to potentiate the effect of NMT1 loss of function upon viability/growth in the CCLE cancer cell lines examined.

Figure 5. Visualisation of NMT1 CRISPR scores by NMT2 gene expression clusters for lung cancer cell lines. Each dot represents a lung cancer cell line. The NMT1 CRISPR score is shown on the y-axis, NMT2 gene expression clusters are shown on the x-axis. Lower CRISPR scores represent greater depletion with NMT1 loss of function and so correspond to a larger effect upon cell viability/growth. The NMT1 CRISPR scores are lower in the first NMT2 expression cluster (left), relative to the second expression cluster (right). Therefore, low expression of NMT2 appears to potentiate the effect of NMT1 loss of function upon viability/growth in the lung cancer cell lines shown.

Enrichment of Mutations in Gene Expression Clusters

SynLeGG displays mutations that are enriched in mRNA expression clusters, providing insight into the genetic background relevant to candidate 'Achilles Heel' relationships suggested by analysis of the CRISPR screen and transcriptome data (Figure 6). A chi-squared test was performed for each gene expression / CRISPR gene pair to derive p-values for the enrichment of mutation classes across the gene expression clusters. Q-values were produced using Benjamini-Hochberg false discovery rate correction [Benjamini & Hochberg 1995].

Figure 6. Visualisation of NMT1 CRISPR scores by NMT2 gene expression clusters for lung cancer cell lines. Each dot represents a lung cancer cell line. The NMT1 CRISPR score is shown on the y-axis, NMT2 gene expression clusters are shown on the x-axis. The colour of each dot represents mutational status, as shown in the key. Cell lines with wild-type (WT) CACNA1B are coloured blue. There is strong enrichment of CACNA1B mutations in the NMT2 high expression group; moreover, cell lines with these mutations and high NMT2 expression are relatively unaffected by NMT1 loss of function in the CRISPR screen. These data are consistent with elevated NMT2 expression as a mechanism to compensate for loss of NMT1 function due to targeting by CRISPR in a CACNA1B mutant background. Similarly, cell lines with low NMT2 expression appear vulnerable to NMT1 loss.

Filtering with Additional Evidence Sources

In addition to statistical comparison of CRISPR scores across mRNA expresssion modes, <SynLeGG integrates further sources of evidence that support gene functional similarity, specifically:
- Overlapping Gene Ontology annotations [Carlson M et al. 2019, Yu G et al. 2010].
- Physical protein-protein interactions (PPI's) from BioGRID [Stark C et al. 2006].
- Paralogous gene pairs; defined by Ensembl and within human only [Zerbino D et al. 2018].

The check-boxes in the 'Optional Results Filters' panel limit the display of results to require the above additional evidence. When multiple checkboxes are ticked, the results shown combine the filters using a logical AND operator. For example selecting both the 'PPI' and 'Gene Ontology' checkboxes displays results where the gene pairs have annotated PPIs and overlapping Gene Ontology terms in addition to passing the statistical filters outlined above.

Plot Types

This section gives a succinct summary for each of the plot types available within SynLeGG.

The 'Integrated' plot shows a boxplot of CRISPR dependency scores. Please see Figure 17, below.


Figure 17. The 'Integrated' Plot. The y axis shows the CRISPR score of the selected CRISPR gene and each box corresponds to the gene expression cluster of the selected mRNA gene. The cell lines are is coloured according to tissue type by default, please see the key for details.
The distribution for the selected mRNA gene is shown in the 'mrna_only' plot. Please see Figure 18, below.


Figure 18. The 'mRNA_only' Plot. The x-axis corresponds to the log2 mRNA expression value and each curve represents a MultiSEp gene expression cluster, the allocation of colourings to cluster identify is shown in the key.
Selecting 'crispr_only' shows a waterfall plot of CRISPR scores for the selected CRISPR gene in the relevant cell lines. Please see Figure 19, below.


Figure 19. The 'Crispr_only' Plot. The bars are ordered from low to high CRISPR score and coloured according to MultiSEp gene expression cluster, the allocation of colourings to cluster identify is given in the key. Lower CRISPR score is associated with a greater effect on cell viability/growth.
The 'Protein' plot shows boxplots of protein concentration values. Please see Figure 20, below.


Figure 20. The 'Protein' Plot. The y axis indicates log2 protein concentration of the selected mRNA genes and each box corresponds to the expression cluster of the selected mRNA gene. The plot is coloured according to MultiSEp gene expression cluster by default, the allocation of colourings to cluster identify is shown in the key.
The 'Cluster' plot shows boxplots of mRNA expression values for the clusters identified by MultiSEp. Please see Figure 21, below.


Figure 21. The 'Cluster' Plot. The y axis provides the log2 expression of the selected mRNA genes and each box corresponds to the expression cluster of the selected mRNA gene. The cell lines are coloured according to mutation class, which is detailed in the key.

The 'Cluster_Type' plot shows nine boxplots of mRNA expression values, one for each mutation class. Please see Figure 22, below.

Figure 22. The 'Cluster_Type' Plot. The y axis shows the log2 gene expression of the selected mRNA gene and each box corresponds to an expression cluster for the selected mRNA gene. The cell lines are coloured according to mutation class, which is shown in the key.

All of the plotting functions in SynLeGG are written in R and implement the ggplot2 package [Wickham H et al. 2016].

Analysis of CRISPR and Gene Expression Data

The CRISPR tab provides integrated gene expression and CRISPR scores for investigation of candidate 'Achilles-Heel' relationships. The 'Results' table (Figure 7) displays gene pairs that pass the MultiSEp thresholds; by default, rows are ordered by qvalue. Results may be filtered by clicking on any of the PPI, Paralogue and Gene Ontology checkboxes. For example, if the PPI box is checked then only mRNA/CRISPR gene pairs that share a physical protein-protein interaction in BioGRID will be returned. If both the 'PPI' and 'Paralogue' boxes are checked then only within-human paralogues with evidence for a physical protein interaction are displayed in the 'Results' table. The 'Search' bar above the Results table allows querying by gene name; there are also individual search boxes for the columns: mRNA_gene, crispr_gene, and cluster. The 'clusters' column refers to the number of gene expression clusters identified by MultiSEp and requires integer values between 2 and 5. For example, entering a value of 3 in the 'clusters' search box will return genes with three expression clusters. Clicking on the 'U' next to each gene name navigates to the relevant UniProt page (opens in a new tab). Clicking the download icon underneath the 'Results' table exports the contents as a .csv file.

Figure 7. Selecting a gene pair in the CRISPR tab. Gene pairs that have dependencies identified by MultiSEp are shown in the Results panel, the gene pair EAF2, EAF1 is selected (blue row). The 'Details' panel shows statistical and Gene Ontology information for the selected gene pairing (EAF2, EAF1).

Clicking on a row in the 'Results' table populates two additional panels; the 'Details' panel (Figure 7) and the 'Plot' panel (Figure 8). The 'Details' panel provides summary statistics and Gene Ontology information, if available, for the selected gene pair. The following statistical information is provided:
- 'Cell Line Number: Cluster X': The number of cell lines in gene expression cluster 'X' (where X has value between 1 and 5).
- 'Average CRISPR Score: Cluster X': The average CRISPR score for each gene expression cluster. Non-essential genes have a median value of 0, essential genes have a median value of -1. Therefore, lower values indicate a stronger effect for the CRISPR gene upon the cell lines in the given expression cluster.
- 'CRISPR Score Fold Change (Cluster X - Cluster Y): The fold-change in CRISPR dependency score between neighbouring clusters (for example, cluster 2 - cluster 1).
- 'CRISPR Score P Value (Cluster X - Cluster Y): The p value relating to the change in CRISPR dependency score between neighbouring clusters (for example, cluster 2 - cluster 1).
- 'CRISPR Score Q Value (Cluster X - Cluster Y): The q value relating to the change in CRISPR dependency score between neighbouring clusters (for example, cluster 2 - cluster 1).

The Gene Ontology Details table has two columns:
- 'Identifier': The identifiers for any overlapping Gene Ontology terms, all three branches are considered (i.e. Molecular Function, Biological Process and Cellular Component).
- 'Name': The name assigned to overlapping GO terms.
Clicking the download icon beneath the 'Gene Ontology Details' table exports all of the information in the 'Details' panel to a .csv file (for the selected gene pair).

The 'Plot' panel shows MultiSEp gene expression clusters and, by default, CRISPR scores from CERES for the selected gene pair ('Integrated' plot type). Several plot types are available from the 'Select Plot Type' drop down menu. The 'crispr_only' plot visualises the CRISPR scores for the 'crispr_gene' selected in the 'Results' panel. The 'mRNA_only' plot visualises the MultiSEp gene expression clusters for the 'mrna_gene' selected in the 'Results' panel. The 'protein' plot visualises protein concentrations from mass spectrometry, where available, for the 'mrna_gene' selected in the 'Results' panel. Clicking the download icon beneath the panel exports the data shown in the plot in .csv format.

Figure 8. The Plot Panel Provides Visualisation of CRISPR scores and Gene Expression Clusters. The Plot type menu is shown at the top of the image, and the 'Integrated' plot type is selected. EAF2 gene expression clusters from MultiSEp are on the x-axis and CRISPR scores are given on the y-axis. The dots represent cell lines and are coloured by tissue of origin, which is detailed in the key. Overall, cell lines in the high EAF2 expression cluster are robust to EAF1 loss of function. On the other hand, many cell lines in the EAF2 low expression cluster have a substantial loss of viability/growth when EAF1 is disrupted, indicated by low CRISPR score values. The median CRISPR score for non-essential genes is 0, the median CRISPR score for essential genes is -1. Please note that plots are generated dynamically and there may be some variability in the x coordinates of individual cell lines that fall within the same mRNA expression cluster.

A relatively small proportion of the exome is covered by the available mass spectrometry proteomics data, compared to transcriptome (RNA-seq) data. However, proteomic information is particularly useful in drug discovery applications because proteins form the molecular machines that ensure cellular function, and are the target of the vast majority of drugs. Protein mass-spectrometry data [Nusinow D et al. 2020] is accessible within the CRISPR tab in the drop-down menu of the 'Plot' section (select 'protein' in the menu). The protein concentration values are visualised in the gene expression clusters from MultiSEp analysis of the corresponding mRNA. Figure 9 shows protein expression for NMT2 gene expression clusters.

Figure 9. Visualisation of NMT2 protein concentrations for MultiSEp gene expression clusters. The y-axis shows protein concentration values, and mRNA expression clusters are given on the x-axis. All of the CCLE cell lines are displayed where mass spectrometry proteomic data is available for NMT2. The high mRNA expression cluster also has higher median NMT2 protein concentration, although some cell lines with a low NMT2 mRNA value have relatively high NMT2 protein concentration and vice versa. Please note that plots are generated dynamically and there may be some variability in the x coordinates of individual cell lines within that fall within the same mRNA expression cluster.

Exploration of gene expression and CRISPR score relationships for individual tissues is available by using the 'Tissue Type' tab. A gene pair must be selected in the 'Results' panel of the 'All Tissue' tab (Figure 7) before navigating to 'Tissue Type' (Figure 10). The 'Tissue Breakdown' table displays individual tissue types for the gene pair selected within the 'All Tissue' tab. Text entered into the 'Search' bar applies a filter to select tissues with matching names. Selecting a row in the 'Tissue Breakdown' table populates the 'Details' and 'Tissue Specific Plot' (Figure 11) panels with information for the tissue of interest. The 'Details' panel (Figure 10) provides summary statistics and includes a 'Top Mutations' table that lists genes for the selected tissue with significantly different mutational profiles between the MultiSEp gene expression clusters. By default, the 'Top Mutations' table is ordered by p-value. The following statistical information is provided:
- 'Cell Line Number: Cluster X': The number of cell lines for the selected tissue in gene expression cluster 'X' (where X has value between 1 and 5).
- 'Average CRISPR Score: Cluster X': The average CRISPR score for each gene expression cluster in the selected tissue. Non-essential genes have a median value of 0, essential genes have a median value of -1. Therefore, lower values indicate a stronger effect for the CRISPR gene upon the cell lines in the given expression cluster.
- 'CRISPR Score Fold Change (Cluster X - Cluster Y): The fold-change in CRISPR dependency score between neighbouring clusters (for example, cluster 2 - cluster 1) in the selected tissue:
- 'CRISPR Score P Value (Cluster X - Cluster Y): The fold-change in CRISPR dependency score between neighbouring clusters (for example, cluster 2 - cluster 1) in the selected tissue.

The 'Top Mutations' table has three columns:
- 'mutation_gene': The gene for which mutations were evaluated across the MultiSEp gene expression clusters.
- 'tissue': The selected tissue.
- 'pvalue': Chi-squared test p-value for the enrichment of mutation classes for the mutation_gene across the MultiSEp gene expression clusters.

Selecting a row in the 'Top Mutations' table displays mutations for the mutation_gene in the 'Tissue Specific Plot' panel, if the 'Integrated' plot type is selected (Figure 12).

Figure 10. Tissue-specific Analysis in the 'Tissue Type' Tab. Selecting the 'Tissue Type' panel displays the 'Tissue Breakdown' table for the gene pair that was selected in the 'All Tissue' tab; EAF2, EAF1 was selected in this example. The row for Lung Cancer is selected (blue) in the 'Tissue Breakdown' table, relevant summary statistics and mutations are displayed in the 'Details' panel (right), including the 'Top Mutations' table (bottom right).

The 'Tissue Specific Plot' panel is shown in Figure 11, which gives MultiSEp gene expression clusters and, by default, CRISPR scores derived from CERES for the selected gene pair and tissue ('Integrated' plot type). Several plot types are available from the 'Select Plot Type' drop down menu. The 'crispr_only' plot visualises the CRISPR scores for the 'crispr_gene' and selected tissue from the 'Tissue Breakdown' panel. The 'mRNA_only' plot visualises the MultiSEp gene expression clusters for the 'mrna_gene' and selected tissue in the 'Tissue Breakdown' panel. The 'protein' plot visualises protein concentrations from mass spectrometry, where available, for the 'mrna_gene' and selected tissue in the 'Tissue Breakdown' panel. Clicking the download icon beneath the panel exports the data shown in the displayed plot in .csv format.

Figure 11. The Tissue Specific Plot Panel Provides Visualisation of CRISPR scores and Gene Expression Clusters. The Plot type menu is shown at the top of the image, and the 'Integrated' plot type is selected. EAF2 gene expression clusters from MultiSEp are on the x-axis and CRISPR scores are given on the y-axis. The dots represent lung cancer cell lines. Overall, cell lines in the high EAF2 expression cluster are robust to EAF1 loss of function. On the other hand, many cell lines in the EAF2 low expression cluster have a substantial loss of viability/growth when EAF1 is disrupted, indicated by low CRISPR score values. The median CRISPR score for non-essential genes is 0, the median CRISPR score for essential genes is -1. Please note that plots are generated dynamically and there may be some variability in the x coordinates of individual cell lines within that fall within the same mRNA expression cluster.

Figure 12. Integrated Visualisation of CRISPR Scores, Gene Expression and Mutations in the Tissue Specific Plot Panel. The Plot type menu is shown at the top of the image, and the 'Integrated' plot type is selected. EAF2 gene expression clusters from MultiSEp are on the x-axis and CRISPR scores are given on the y-axis. The dots represent lung cancer cell lines, coloured by mutations in the gene selected in the 'Top Mutations' panel. Please note that plots are generated dynamically and there may be some variability in the x coordinates of individual cell lines within that fall within the same mRNA expression cluster.

Terms of Use

SynLeGG is free of charge for all to use (including for commercial use). We make every effort to ensure that the software is free from errors, however all usage of SynLeGG implies acceptance of the disclaimer notice. Usage of SynLeGG also implies acceptance of these Terms of Use and the Privacy & Cookie Policy.
If you use SynLeGG in any published academic or commercial materials, please cite:

Wappett M, Harris A, Lubbock ALR, Lobb I, McDade S, Overton IM. SynLeGG: finding ‘Achilles Heel’ relationships in large CRISPR screens using multi-modal mRNA expression. Manuscript in preparation. URL: www.overton-lab.uk/synlegg

DepMap, Broad (2020): DepMap 20Q3 Public. figshare. Dataset doi:10.6084/m9.figshare.12931238.v1.

Robin M. Meyers, Jordan G. Bryan, James M. McFarland, Barbara A. Weir, ... David E. Root, William C. Hahn, Aviad Tsherniak. Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells. Nature Genetics 2017 October 49:1779–1784. doi:10.1038/ng.3984

Dempster, J. M., Rossen, J., Kazachkova, M., Pan, J., Kugener, G., Root, D. E., & Tsherniak, A. (2019). Extracting Biological Insights from the Project Achilles Genome-Scale CRISPR Screens in Cancer Cell Lines. BioRxiv, 720243.

Mahmoud Ghandi, Franklin W. Huang, Judit Jané-Valbuena, Gregory V. Kryukov, ... Todd R. Golub, Levi A. Garraway & William R. Sellers. 2019. Next-generation characterization of the Cancer Cell Line Encyclopedia. Nature 569, 503–508 (2019).

Also please see the depmap terms and the CC Attribution 4.0 (CC BY) license. We hope you find SynLeGG a useful research tool. We are always keen to hear feedback and comments. Please contact us with your views, they are much appreciated.

Disclaimer Notice

DISCLAIMER OF LIABILITIES AND WARRANTIES

THE SYNLEGG WEBSITE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS AND/OR QUEEN'S UNIVERSITY BELFAST AND/OR ALMAC DIAGNOSTICS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SYNLEGG WEBSITE OR THE USE OR OTHER DEALINGS IN THE SYNLEGG WEBSITE.

QUEEN'S UNIVERSITY BELFAST AND/OR ALMAC DIAGNOSTICS MAKE NO WARRANTIES ABOUT THE ACCURACY, RELIABILITY, COMPLETENESS, OR TIMELINESS OF THE MATERIAL, SERVICES, SOFTWARE, TEXT, GRAPHICS AND LINKS. DESCRIPTION OF OR REFERENCES TO PRODUCTS OR PUBLICATIONS DOES NOT IMPLY ENDORSEMENT OF THAT PRODUCT OR PUBLICATION.

BY USING THE SYNLEGG WEBSITE YOU AGREE TO ASSUME ALL RISKS ASSOCIATED WITH YOUR USE OR TRANSFER OF ANY AND ALL INFORMATION CONTAINED ON THIS SITE AND TO HOLD THE THE AUTHORS AND/OR QUEEN'S UNIVERSITY BELFAST AND/OR ALMAC DIAGNOSTICS HARMLESS FROM ANY CLAIMS RELATING TO CONTENT OR INFORMATION IN EXCHANGE FOR YOUR USE OF THE SITE.

THE AUTHORS AND/OR QUEEN'S UNIVERSITY BELFAST AND/OR ALMAC DIAGNOSTICS DOES NOT WARRANT THE SYNLEGG WILL OPERATE ERROR-FREE OR THAT THIS SITE AND ITS SERVER ARE FREE OF COMPUTER VIRUSES OR OTHER HARMFUL MECHANISMS. IF YOUR USE OR TRANSFER OF THIS SITE OR THE MATERIALS RESULTS IN THE NEED FOR SERVICING OR REPLACING EQUIPMENT OR DATA, THE AUTHORS AND/OR QUEEN'S UNIVERSITY BELFAST AND/OR ALMAC DIAGNOSTICS IS NOT RESPONSIBLE FOR THOSE COSTS.

INDEMNITY

YOU AGREE TO DEFEND, INDEMNIFY AND HOLD HARMLESS THE QUEEN'S UNIVERSITY BELFAST AND/OR ALMAC DIAGNOSTICS, ITS OFFICERS, DIRECTORS, EMPLOYEES AND AGENTS FROM AND AGAINST ANY CLAIMS ACTIONS OR DEMANDS, (INCLUDING WITHOUT LIMITATION ALL LEGAL AND ACCOUNTING FEES) WHICH MAY ARISE DUE TO YOUR USE OR TRANSFER OF THE SYNLEGG WEBSITE, ITS CONTENTS, OR INFORMATION THEREOF.

Privacy and Cookie Policy

By using the SynLeGG website, you consent to the collection, retention and use of your personal information in accordance with the terms of this policy.

Information we collect

- The supplied name and email address if you submit a query using our contact form
- Information on how you use the website and any other information you post, email or otherwise send to us.
- Personally identifiable information will not be distributed to third parties unless required for the purposes of law enforcement.

We may gather data about users' browsing actions and patterns in order to inform improvements to the website. We do not track the identities of genes or gene pairs entered or accessed by users, nor do we track the frequency of access to any gene or gene pair.

Security

We employ security measures to protect your information from access by unauthorised persons and against unlawful processing, accidental loss, destruction or damage.

Unfortunately, the transmission of information via the internet is never completely secure. Although we will do our best to protect your personal data, we cannot guarantee the security of your personal or other data transmitted to our website. Any transmission is at your own risk.

Once we have received your information, we will use strict procedures and security features to prevent unauthorised access.

Cookies on SynLeGG

EU law
The EU law on cookies demands that website users are given the opportunity to understand how cookies are used on websites and consent to cookies being stored on their computer.

What are cookies?
A cookie is a small text file, typically containing letters and numbers, downloaded to your computer when you access websites.

When you visit a website that uses cookies for the first time, a cookie is downloaded onto your computer. The next time you visit that website, your computer checks to see if it has a cookie that is relevant and sends the information contained in that cookie back to the website. The website then notes that you have been there before, and in some cases, tailors what pops up on screen to take account of that fact. They also might record how long you spend on each page on a site, what links you click, even your preferences for page layouts and colour schemes.

Generally, the role of cookies is beneficial, making your interaction with frequently-visited sites smoother with no extra effort on your part. Without cookies, online shopping would be much harder. Without cookies, some websites will become less interactive with the cookie option turned off.

First and third party cookies
This refers to the website placing the cookie. First party cookies are cookies set by the website you are visiting. Third party cookies are set by another website; the website you are visiting may have advertising on the page and this other website will be able to set a cookie on your computer. Third party cookies on the main web browsers allow third party cookies by default. Changing the settings on your browsers can prevent this.

Our use of cookies
We do not use any third party cookies. However, use of SynLeGG places a session cookie on your computer. Use of this website implies that you are happy to accept cookies from SynLeGG.

Citing SynLeGG

To cite SynLeGG, please reference the publication below:

Wappett M, Harris A, Lubbock ALR, Lobb I, McDade S, Overton IM. SynLeGG: finding ‘Achilles Heel’ relationships in large CRISPR screens using multi-modal mRNA expression. Nucleic Acids Research, Volume 49, Issue W1, 2 July 2021, Pages W613–W618, DOI: https://doi.org/10.1093/nar/gkab338