Determining the structure of proteins can be a long expensive task. However, using new applications of deep learning ai it is possible to formulate a protein structure using its sequence. Through many supporting applications the final goal is becoming more attainable.
Bioinformatics study on protein modeling Read More »
Proteins, a class of macromolecules essential to biological processes, are characterized by their structure which directly correlates to functionality. Residues are considered the building blocks of proteins and studying their positions has been crucial in understanding the role a protein plays in a biological system. This research project investigates different measures of residue depth through calculations of RSA, DPX, Residue Depth, HalfSpace Depth, and L1 depth. The L1 depth function surpasses other residue structure predictors in that it can obtain positions of residues buried under the protein surface, or nested in pockets. In addition, the function uses O(N) time complexity, making it much more efficient than the HalfSpace predictor. Values of depth have strong correlations with properties such as physiochemical propensities, flexibility and polarity. Data was obtained from CULLPDB, Phospho3D, and PLB datasets. After reading in necessary information into an IDE, residue depth of each protein in the datasets and the correlation coefficients between the means of depth values of amino acids and hydrophobicity index of amino acids were calculated through an algorithm written in C++. These results showed that the means of depth values were strongly relative with hydrophobicity of amino acids. In addition, the secondary structures of the residues are also associated with L1 depth values, as residues in sheets are deeper under the surface in comparison to residues in coils and helices.
Residue Depth Computation Read More »
Proteins, a class of macromolecules essential to biological processes, are characterized by their structure which directly correlates to functionality. Residues are considered the building blocks of proteins and studying their positions has been crucial in understanding the role a protein plays in a biological system. This research project investigates different measures of residue depth through calculations of RSA, DPX, Residue Depth, HalfSpace Depth, and L1 depth. The L1 depth function surpasses other residue structure predictors in that it can obtain positions of residues buried under the protein surface, or nested in pockets. In addition, the function uses O(N) time complexity, making it much more efficient than the HalfSpace predictor. Values of depth have strong correlations with properties such as physiochemical propensities, flexibility and polarity. Data was obtained from CULLPDB, Phospho3D, and PLB datasets. After reading in necessary information into an IDE, residue depth of each protein in the datasets and the correlation coefficients between the means of depth values of amino acids and hydrophobicity index of amino acids were calculated through an algorithm written in C++. These results showed that the means of depth values were strongly relative with hydrophobicity of amino acids. In addition, the secondary structures of the residues are also associated with L1 depth values, as residues in sheets are deeper under the surface in comparison to residues in coils and helices.
Residue Depth Computation Project Read More »
Protein-protein interactions (PPIs) are physical contacts between two or more protein molecules. The field of structural homology-based PPI prediction, has generally accepted that the interface region of the complex structure matters more than the global structure of proteins in predicting PPIs, although our preliminary data suggests otherwise. A study was undertaken to investigate the influence of the protein interface in predicting PPIs, using a tweaked version of TMalign. This edited version of the algorithm allows a weighting factor to be applied to specific residues, the interface residues, in these experiments. This study consists of experiments using different weighting values, each of which will be done on three sets of query data. The three data sets are: 1. pairs of proteins that are known to have positive interactions, 2. pairs of proteins that are said to not interact with one another, and 3. a randomized data set. The latter two will allow us to confirm the results of non-interactions. Once each query protein is aligned to the structural homologs in the database, the score can be calculated and then all scores will be sorted to determine the cutoff for classifying predicted positive interactions. The results allow us to determine how impactful the interface is in predicting whether two given proteins will interact or not. Additionally, it will either confirm or deny the long standing assumption that only the interface matters.
The importance of the interface in predict protein-protein interactions Read More »
The efficient searching of existing genomic markers is essential for expediting the analysis of key biological information and further manipulating the data in bulk to discover patterns. In particular, the Cutaneous Bioinformatics project aims to process genetic mutations, namely simple nucleotide polymorphisms and insertions-deletions, to deduce how certain mutations facilitate certain epidemiological conditions. To efficiently analyze the genomic markers, a C++ program was developed, tested, and published on GitHub that reads the data from a standard tab-separated value text format and inputs the data into a two-way hash map. The program was developed, tested, and documented by a single individual under the supervision and direction of the leaders of the Cutaneous Bioinformatics project. The software takes command-line arguments and can perform two-way lookups between the markers’ rsIDs and their chromosome, position, and allele sequence. The utilization of a hash table is ideal because it allows lookups in both directions to be performed with constant-time complexity, while manual analysis of the original, multi-gigabyte data file requires linear-time complexity for reverse lookups. The hash table creation was successfully implemented, and, upon deliberations with project supervisors, the syntax of user input for allele sequences in the reverse lookup was taken into account when determining which markers with a particular chromosome and position constitute a match. The incorporation of additional markers and data across other biological disciplines can further augment the ability of researchers to quickly analyze the data and perhaps lead to genetic discoveries.
Background: Rates of sexually transmitted infections (STIs) in the United States are rising dramatically. EDs serve as the primary source of care for many patients diagnosed with STIs. Expedited Partner Therapy (EPT), an evidence-based practice for treating partners of patients with STIs without an exam, can decrease rates of STI reinfection and increase the number of partners treated. Despite the potential benefits of EPT, its use in EDs is poorly understood. Objective: This study characterizes the knowledge, beliefs, interest, and practice of EPT among academic ED medical directors through a cross-sectional survey. We also explore barriers and facilitators to EPT implementation through semi-structured interviews of medical directors and key informants conducted using an interview guide informed by the Consolidated Framework for Implementation Science. Interviews were recorded, transcribed, and iteratively coded, and qualitative thematic analysis was used to identify themes and subthemes by consensus. Results: A total of 48 out of 70 medical directors responded. There was a high awareness of EPT (73%), but only 38% knew how to prescribe EPT, and even fewer (19%) had actually implemented it. Most respondents supported EPT (79%), and were more likely to support EPT if they were already aware of it (89% vs 54%). 41% of those who had not implemented EPT thought that it would be feasible, and 56% thought that departmental support would be likely. Many participants cited medicolegal barriers and lack of clarity regarding the ED’s role in STI care as obstacles to EPT adoption. Other participants cited a perceived benefit of EPT use was the prevention of untreated STIs, particularly among a vulnerable population. Conclusion: Our findings show relatively low use but high interest in EPT in academic EDs, and show the need for increased awareness to implement this public health intervention in the ED setting.
A National Study of Expedited Partner Therapy in Academic Emergency Departments Read More »
Background: Rates of sexually transmitted infections (STIs) in the United States are rising dramatically. EDs serve as the primary source of care for many patients diagnosed with STIs. Expedited Partner Therapy (EPT), an evidence-based practice for treating partners of patients with STIs without an exam, can decrease rates of STI reinfection and increase the number of partners treated. Despite the potential benefits of EPT, its use in EDs is poorly understood. Objective: This study characterizes the knowledge, beliefs, interest, and practice of EPT among academic ED medical directors through a cross-sectional survey. We also explore barriers and facilitators to EPT implementation through semi-structured interviews of medical directors and key informants conducted using an interview guide informed by the Consolidated Framework for Implementation Science. Interviews were recorded, transcribed, and iteratively coded, and qualitative thematic analysis was used to identify themes and subthemes by consensus. Results: A total of 48 out of 70 medical directors responded. There was a high awareness of EPT (73%), but only 38% knew how to prescribe EPT, and even fewer (19%) had actually implemented it. Most respondents supported EPT (79%) and were more likely to support EPT if they were already aware of it (89% vs 54%). 41% of those who had not implemented EPT thought that it would be feasible, and 56% thought that departmental support would be likely. Many participants cited medicolegal barriers and lack of clarity regarding the ED’s role in STI care as obstacles to EPT adoption. Other participants cited a perceived benefit of EPT use was the prevention of untreated STIs, particularly among a vulnerable population. Conclusion: Our findings show relatively low use but high interest in EPT in academic EDs, and show the need for increased awareness to implement this public health intervention in the ED setting.
Increasing access to STD treatment in the Emergency Department through EPT Read More »
Awareness and adoption of machine learning (ML) are growing in the medical field, where ML models can support clinical decision-making when linked to interventions. The process of developing and evaluating ML models is an important precursor to their deployment in a health setting. This process can be made transparent, reliable, and reproducible through the development and dissemination of open-source software packages. My project focuses on the development and dissemination of runway, an R package that evaluates and compares prediction models using statistics and plots. My primary aim is to understand the lifecycle of an R package””from creation to completion to dissemination””by working through each of the steps in the runway package. Fixing issues submitted to the GitHub page for runway and using Git to update changes to the package were important steps in the process of preparing for submission to the Comprehensive R Archive Network. Runway is in the final stages of development. The final product will be submitted to CRAN, making it widely available and easy to install for researchers around the world.
Development of R Packages for the Machine Learning for Learning Health Systems Lab Read More »
Transposable elements (TEs), DNA sequences that can change their position within the genome, can result in mutations associated with somatic and heritable diseases. TEs are repetitive in the genome and are therefore hard to map. In this study, we determined definitive mapping of one set of TEs, called Long Interspersed Nuclear Elements 1 (L1), through the utilization of recent technology, specifically CRISPR-Cas9 and nanopore sequencing. However, the cutting precision of Cas9 must be deduced for efficient mapping of these elements. This study aims to better understand the cutting preferences of Cas9 in the context of the L1 sequence by exploring how Python can map transposable elements in the genome. The biological analysis and mapping of these reads through the use of Biopython will result in a better understanding of the cutting preferences of Cas9 and its efficiency. Through an RNA guide, Cas9 targets and cuts at a specific region of DNA, where nanopore sequencing will then read the L1 retrotransposons. This data will be processed through the use of Biopython’s modules. A “for” loop was used to: extract individual reads contained within the input data files; perform a local alignment of each read to L1; and obtain the positions of the alignments with respect to L1. We also accounted for reads that align to the reverse complement of the L1 sequence by comparing the two scores obtained from the local alignments. Our data demonstrates that the median starting alignment position is consistent between alignments of the reads to L1 and its reverse complement. Thus, we conclude that Cas9 cutting occurs approximately 5900 bp downstream in L1. However, further analysis of the location of PAM sites and the RNA guide is necessary to confirm Cas9’s efficiency and function.
Transposable elements (TEs), DNA sequences that can change their position within the genome, can result in mutations associated with somatic and heritable diseases. TEs are repetitive in the genome and are therefore hard to map. In this study, we determined definitive mapping of one set of TEs, called Long Interspersed Nuclear Elements 1 (L1), through the utilization of recent technology, specifically CRISPR-Cas9 and nanopore sequencing. However, the cutting precision of Cas9 must be deduced for efficient mapping of these elements. This study aims to better understand the cutting preferences of Cas9 in the context of the L1 sequence by exploring how Python can map transposable elements in the genome. The biological analysis and mapping of these reads through the use of Biopython will result in a better understanding of the cutting preferences of Cas9 and its efficiency.
