BOL: Related items

Next generation sequencing in R or bioconductor environment

John Parker — Mon, 02 Jun 2014 18:03:09 -0500

There are many R software and bioconductor packages for NGS data analysis, some of them are as follows

Biostrings

The Biostrings package from Bioconductor provides an advanced environment for efficient sequence management and analysis in R. It contains many speed and memory effective string containers, string matching algorithms, and other utilities, for fast manipulation of large sets of biological sequences. The objects and functions provided by Biostrings form the basis for many other sequence analysis packages. Documentation

IRanges Overview

IRanges provides the low-level infrastructure and containers for handling sets of integer ranges within Bioconductor's BioC-Seq domain. Its classes and methods provide support for many more high-level packages like GenomicRanges, ShortRead, Rsamtools, etc. Documentation

GenomicRanges Overview

The GenomicRanges package serves as the foundation for representing genomic locations within the Bioconductor project. It is built upon the IRanges infrastructure and defines three major data containers - GRanges, GRangesList and GappedAlignments - which are supporting other important BioC-Seq packages including ShortRead, Rsamtools, rtracklayer, GenomicFeatures and BSgenome. Compared to the IRanges container, the GRanges/GRangesList classes are more flexible and extensible to store additional information about sequence ranges, such as chromosome identifiers (sequence space), strand information and annotation data. Documentation

Motif Discovery

cosmo

The cosmo package allows to search a set of unaligned DNA sequences for a shared motif that may function as transcription factor binding site. The algorithm extends the popular motif discovery tool MEME (Bailey and Elkan, 1995) in that it allows the search to be supervised by specifying a set of constraints that the motif to be discovered must satisfy. Documentation

BCRANK

BCRANK is a method that takes a ranked list of genomic regions as input and outputs short DNA sequences that are overrepresented in some part of the list. The algorithm was developed for detecting transcription factor (TF) binding sites in a large number of enriched regions from high-throughput ChIP-chip or ChIP-seq experiments, but it can be applied to any ranked list of DNA sequences. Documentation

rGADEM: Documentation

MotIV: Documentation

ShortRead

The ShortRead package provides input, quality control, filtering, parsing, and manipulation functionality for short read sequences produced by high throughput sequencing technologies. While support is provided for many sequencing technologies, this package is primairly focused on Solexa/Illumina reads. Documentation

Rsamtools

Rsamtools provides functions for parsing and inspecting samtools BAM formatted binary alignment data. SAM/BAM is quickly becoming a universal standard alignment format, and is now supported by a wide variety of alignment tools. Documentation

Samtools Website
BWA (Burrows-Wheeler Alignment) Website

Additional tools for SNP analysis:

snpMatrix

BSgenome

BSgenome provides an object oriented infrastructure for interacting with a Biostring based genome sequence. BSgenome packages exist for many common genomes, and can be created to represent custom genomes. See the "How to forge a BSgenome data package" Vignette for instructions to create a new BSgenome package if a prebuilt package does not exist for your organism. Documentation

rtracklayer

rtracklayer provides an interface for exporting annotation feature data to various genome browsers and file formats (such as GFF). See the Small RNA Profiling exercise for an example of using rtracklayer to visualize alignment coverage. Documentation

biomaRt

The biomaRt package, provides an interface to a growing collection of databases implementing the BioMart software suite (http:// www.biomart.org). The package enables online retrieval of large amounts of data in a uniform way without the need to know the underlying database schemas. This data is retrieved automatically via the Internet, so it's recommended that you cache the data locally, or check versions if your code will be adversely affected by updates to these data. Documentation

ChIP-Seq Analysis Packages

Bioconductor provides various packages for analyzing and visualizing ChIP-Seq data. Only a small selection of these packages is introduced here. Additional useful introductions to this topic are: BioC ChIP-seq Case Study and BioC ChIP-Seq.

chipseq

The chipseq package combines a variety of HT-Seq packages to a pipeline for ChIP-Seq data analysis. Documentation

BayesPeak

BayesPeak is a peak calling package for identifying DNA binding sites of proteins in ChIP-Seq experiments. Its algorithm uses hidden Markov models (HMM) and Bayesian statistical methods. The following sample code introduces the identification of peaks with the BayesPeak package as well as the incorporation of read coverage information obtained by the chipseq package. Documentation [ Publication ]

PICS

The PICS package applies probabilistic inference to aligned-read ChIP-Seq data in order to identify regions bound by transcription factors. PICS identifies enriched regions by modeling local concentrations of directional reads, and uses DNA fragment length prior information to discriminate closely adjacent binding events via a Bayesian hierarchical t-mixture model. The following sample code uses the test data set from the above BayesPeak package in order to compare the results from both methods by identifying their consensus peak set. Documentation [ Publication ]

ChIPpeakAnno

The ChIPpeakAnno package provides. batch annotation of the peaks identified from either ChIP-seq or ChIP-chip experiments. It includes functions to retrieve the sequences around peaks, obtain enriched Gene Ontology (GO) terms, find the nearest gene, exon, miRNA or custom features such as most conserved elements and other transcription factor binding sites supplied by users. The package leverages the biomaRt, IRanges, Biostrings, BSgenome, GO.db, multtest and stat packages. Documentation

Additional ChIP-Seq Packages

DiffBind: Documentation

MOSAICS: Documentation

iSeq: Documentation

ChIPseqR: Documentation

ChiPsim: Documentation

CSAR: Documentation

ChIP-Seq Pipeline: PICS, rGADEM and MotIV (developer web site)

SPP: ChIP-seq processing pipeline

SPP Tutorial

MACS

SIPeS

RNA-Seq Analysis

Counting Reads that Overlap with Annotation Ranges

The GenomicRanges package provides support for importing into R short read alignment data in BAM format (via Rsamtools) and associating them with genomic feature ranges, such as exons or genes. This way one can quantify the number of reads aligning to annotated genomic regions. The package defines general purpose containers for storing genomic intervals as well as more specialized containers for storing alignments against a reference genome. The two main functions for read counting provided by this infrastructure are countOverlaps and summarizeOverlaps. For their proper usage, it is important to read the corresponding PDF manual. Documentation

Differential Gene Expression Analysis with DESeq

The DESeq package contains functions to call differentially expressed genes (DEGs) in count tables based on a model using the negative binomial distribution. It expects as input a data frame with the raw read counts per region/gene of interest (rows) for each test sample (columns). Such a count table can be imported into R or generated from BAM alignment files using the countOverlaps function as introduced above. Documentation

Differential Gene Expression Analysis with edgeR

The edgeR package uses empirical Bayes estimation and exact tests based on the negative binomial distribution to call differentially expressed genes (DEGs) in count data.

Documentation

A variety of additional R packages are available for normalizing RNA-Seq read count data and identifying differentially expressed genes (DEG):

easyRNASeq (simplifies read counting per genome feature)

DEXSeq (Inference of differential exon usage); parathyroidSE explains how to generate exon read counts in R

DEGseq

baySeq (also see: segmentSeq)

Genominator (Bullard et al. 2010)

Detection of Alternative Splice Junctions

Another utility of RNA-Seq experiments is the analysis of splice junctions. The following software suggestions provide this utility:

ERANGE
TopHat

SpliceMap

SplitSeek

DNA-Methylation Data Analysis

methylPipe
bsseq
BiSeq
Much more under BiocViews

HT-Seq Data Visualization

ggbio: ggplot2 extension for genomics data (online manual) Gviz: Plotting data and annotation information along genomic coordinates HilbertVis: Hilbert genome plots

GenomeGraphs: Plotting genomic information from Ensembl

TileQC: Flow Cell Quality Visualization

rtracklayer: R interface to genome browsers

genoPlotR: Plotting maps of genes and genomes

Genominator: Tools for storing, accessing, analyzing and visualizing genomic data.

To install all packages

source("http://bioconductor.org/biocLite.R")
biocLite()
biocLite(c("ShortRead", "Biostrings", "IRanges", "BSgenome", "rtracklayer", "biomaRt", "chipseq", "ChIPpeakAnno", "Rsamtools", "BayesPeak", "PICS", "GenomicRanges", "DESeq", "edgeR", "leeBamViews", "GenomicFeatures", "BSgenome.Celegans.UCSC.ce2"))

Bioinformatics algorithms tutorials

John Parker — Tue, 24 Jun 2014 00:10:45 -0500

Useful bioinformatics tutorial, such as

De Bruijn Graphs for NGS Assembly
Algorithms for PacBio Reads
Software and Hardware Concepts for Bioinformatics
Finding us in Homolog.us (Search Algorithms)
NGS Genome and RNAseq Assembly - a Hands on Primer
Introduction to PERL, Python, R and C/C++ for Bioinformatics

Address of the bookmark: http://www.homolog.us/Tutorials/

RStudio

Jitendra Narayan — Sat, 27 Dec 2014 06:50:58 -0600

RStudio IDE is a powerful and productive user interface for R. It’s free and open source, and works great on Windows, Mac, and Linux.

The developers and expert trainers are the authors of several popular R packages, including ggplot2, plyr, lubridate, and others.

More at http://www.rstudio.com/

http://www.rstudio.com/products/RStudio/

Address of the bookmark: http://www.rstudio.com/

seqloc 0.6

Gudiya Pal — Sun, 28 Dec 2014 12:51:29 -0600

The Bio.SeqLoc modules in seqloc are designed to represent positions and locations (ranges of positions) on sequences, particularly nucleotide sequences. My original motivation for writing these packages was handing the locations of genes in eukaryotic genomes.

Handle sequence locations for bioinformatics http://www.ingolia-lab.org/seqloc-tutorial.html

Address of the bookmark: http://www.stackage.org/snapshot/nightly-2014-12-28/package/seqloc-0.6

Bioinformatics Scripts

Jit — Thu, 22 Jan 2015 22:29:39 -0600

Some of the useful bioinformatics scripts.

For example ... contig-stats.pl is a Perl script that will automatically describe features of a sequence assembly.

http://milkweedgenome.org/?q=scripts

Address of the bookmark: http://milkweedgenome.org/?q=scripts

Bioinformatics PhDs - Cover Letter

Martin Jones — Mon, 16 Feb 2015 12:48:52 -0600

Overview
The reception your cover letter will receive is more varied and unpredictable than the other elements of your application packet. Some readers, especially at large research universities, will skip it entirely, and focus instead on more direct indicators of your academic achievements and potential: the CV, letters, and writing sample. Most often, however, your cover letter works in tandem with your CV, and represents your best opportunity to communicate directly with your target audience prior to an interview.

The cover letter should not simply repeat the information contained in the CV; rather it should elaborate and frame the aspects of your academic training and background that you want search committee members to have uppermost in their minds as they consider your candidacy. At a minimum, it should contain a clear statement of your research and teaching interests, and how your qualifications match the requirements of the job description. Many disciplines have their own conventions and protocols for application materials including the cover letter. Be sure to show a draft of your letter to your chair or the department's placement advisor.

There is no perfect or even preferred style of cover letter save that tone should be that of a confident professional. Departments aren't interested in hiring graduate students (they already have enough of those). They are looking for serious scholars/teachers who will make interesting, congenial, and productive colleagues. While it is probably not a good use of your time to tailor your letter for each opening, you will probably need at least two base versions that emphasize different elements (You would not want to use the same cover letter to apply to Oberlin as UC Irvine). You may want to customize the letters for the three to five jobs most attractive to you.

Your cover letter should not exceed a page and a half or two at the most. No one will read a four-page letter, and your apparent inability to communicate your credentials in a concise manner is not to your advantage. Every line of your letter should serve a demonstrable purpose. Some people have gotten excellent jobs with a brief, boilerplate letter containing no more than a short intro, a paragraph on their research, and a couple of sentences on their teaching. Others use letters that discuss in very specific terms how they, their research, and their teaching would "fit" within the existing department and institutional setting. The choice of style is up to you, and should reflect what makes you feel most comfortable and most positive about your credentials.

There are, however, circumstances where a longer, more annotated version is more likely to be helpful. The issues you should take into account when making this decision are:

    The size of the department/institution
    The extent to which your research is mainstream and its significance readily apparent
    The extent to which your qualifications diverge from those mentioned in the job ad
    The extent to which the institution to which you are applying differs from UC Berkeley and the relative importance of teaching versus research
    Unusual circumstances or career paths

Size matters. The larger the department, the less detailed your letter needs to be. If there are already six people working in your sub-field, they can assess and translate the significance of your work to others in the department. If, however, you will be the only history of science person or the only physical anthropologist in the department, the search committee and others will likely need more help from you to understand the import of your work. Remember, if they are searching for your specialty they are understaffed in your area, and even if there is more than one person in your sub-field she/he may be on leave or otherwise uninvolved in the search.

Similarly, smaller departments are more likely to be concerned than larger ones about the specific courses you are prepared to teach. If there are certain core courses your position needs to cover, make it unambiguously clear that you are prepared to do so even if you haven't taught them already.

As a rule of thumb, the more your profile diverges from the specifics of the job description (or the norm, such that it exists), the more likely you will want to say more about why you are nonetheless a strong candidate. By doing so you can turn a question mark into an advantage. Highlight your strengths, and if you have time perform a little research on the department (see below) so you can explain how you would add to the department in ways that they might otherwise not have considered. Job descriptions are not always etched in stone. If you're very interested in the job, call the department and ask to speak to someone on the search committee. If that's too intimidating, call the departmental assistant or secretary. They can at least tell you if it’s a new position or replacement which in turn can indicate whether they really need someone who can teach The Politics of Southern Africa or if Middle Eastern Politics would do just as well. They may well be willing to sacrifice a course in an area that the ad specified in order to get the extra teaching experience or innovative research that you have to offer, but only if you give them the opportunity and explain what they'd get in return.

If you've taken an unusually long time to finish because you spent three years learning a new language in order to prepare for two years of original research or some other factor that enriches what you have to offer as a teacher or a scholar, let them know about it. Learning a new language, for example, is an indicator of your commitment to serious scholarship. Think of a way to put a positive valence on an aspect of your CV that you believe is likely to raise questions or doubts.

As hard as it may be for not-yet-employed-but-soon-to-be academics like yourselves to believe, there are some individuals and departments that are intimidated by you and your institutional pedigree. There are many fine schools and departments that are convinced that every Berkeley graduate student is only interested in a position at Harvard, Princeton, or (hold your nose) Stanford. If it is a small liberal arts college, they may also assume that you view teaching and advising undergraduates strictly as a necessary evil. Ergo, why interview someone who will either look down their nose at us, and/or desert us for greener pastures as soon as their third book has been published and Yale comes calling. They often don't seem to realize that Berkeley graduate students are sincerely interested in a wide range of academic settings for a variety of reasons personal and professional, and, perhaps more to the point, Yale rarely "comes calling." An even greater burden of proof exists if a quick scan of your CV reveals that your all of your experience in higher education has occurred in a large, elite, research-oriented setting. How can you overcome this potential obstacle especially if you're very interested in a school that you have reason to believe may be suspicious of Berkeley PhDs?

It's probably not advisable to write in a cover letter that they shouldn't be intimidated by lil' old you (i.e., "I'm really not very good; in fact, my advisor says I'm her worst graduate student in twenty years."), but there are ways to indicate a sincere interest in their department. One means is to do a little research, and briefly discuss how you might fit into the department and the institutional community-at-large. There are numerous articles written by faculty about the search process that state explicitly the positive impact it can have on their impression of a candidate to see evidence that that he/she has taken the time to learn something about them. Researching a department and its associated programs/research centers has never been easier. Virtually all colleges and universities are online, and their web addresses can be easily located using Education/Universities link.

A second way to address fears of imminent flight is to provide evidence of ties to the area, the university, or lifestyle.   If you are from the Midwest, have family nearby, or even if you just spent an enjoyable summer there, add a line about it and your desire to return to your cover letter for Purdue or Wayne State. If you did your undergraduate work at a similar type of institution, draw their attention to that link. Don't assume that they will notice where you got your BA on your CV, and make the connection.

If you are applying to small, liberal arts colleges, don't just list the courses you've taught with a line of boilerplate about how important you take your teaching responsibilities to be. These days, a good, small college has a list price of about $30,000 a year, and close contact with people like you is one of their key selling points to parents and potential students. You are expected to be more than a lecturer, and your ability to convey your recognition of that fact and a sincere interest in working closely with students matters. Advising, participating in non-academic activities, watching your students grow and mature inside the classroom and out (and having an influence on that process) can be among the most rewarding aspects of the profession. If you value these broader elements of being a professor, let them know. In particular, if your own experience has been limited to large universities, think about describing an episode where you had a positive impact on the development of a student and the satisfaction you derived from helping him or her. At many small colleges, and large universities (including Berkeley) as well, one of the most significant pedagogical trends is fostering greater student involvement in faculty research. If you can articulate how this might occur for you and their students in a brief but thoughtful manner, it can alleviate some of the concerns occasioned by your Berkeley background.

Your letter typically will have four segments: the introduction, research, teaching, and the closing.

Introduction
Be sure to identify the position you are applying for by rank and sub-field in the first sentence or two. It is not uncommon for departments to be engaged in more than one search in a given year, and large departments may have more than one in your sub-field. If it is an open rank position and you are applying as an assistant professor, in most cases you and your record of accomplishments will not be directly compared with those of more experienced candidates applying for it at the associate or full professor level. Rather, applicants are assessed based on where they are in their career path, and you have every chance of successfully competing against them.

If you have finished, mention it up front. If not, state when you expect to file - no later than June 2003. Many readers will view your optimistic prediction with a skeptical eye, and anything you can say that makes the claim appear more credible (e.g., "I have written and my chair has reviewed four of the six chapters.") can help mitigate their concerns. You can put such a statement in the introduction or wait until you discuss the dissertation itself.

If your discipline holds its annual meeting in the summer or early fall (i.e., before application deadlines) and you had an excellent conference interview, make reference to it in the cover letter including the names of the faculty with whom you spoke. By the time people actually start to read files, months may have passed and even the strongest of impressions can fade. But they can be rehabilitated and revived, especially if you can remind them of a specific strength, ability, or issue that seemed particularly salient during the interview.
Research
Unless you are applying to a school that cares only about your teaching (increasingly rare), a description of your research generally follows next. The challenge here is not simply to describe your research, but to frame it in terms of your sub-field and discipline. The search committee, hasn't lived, eaten and breathed Post-Edwardian Hermeneutics for the past five years the way you have. And they haven't been there every step of the torturous process like your friends and significant others. So forgive them for not immediately recognizing your research at first blush for the path-breaking work that it is.

Departments want to know that in hiring you, they are adding someone who will make a future contribution to the discipline and enhance the reputation of the department. However, they are often ill-equipped to understand exactly how that will be true in your particular case. Many disciplines are sufficiently broad that leading or cutting-edge research in one sub-field is barely intelligible to those in others. In addition, when you became a candidate for the short list, your file will be read by department members outside your specialty, and, oftentimes later on in the process, by individuals outside your discipline as well. You need to describe the forest in which your tree resides, and explain why it matters in terms of the broader trends and issues within your discipline. Obviously, if your work is focused on one of the classic conundrums of your field, much less in the way of providing perspective is necessary than if you are addressing an emergent issue or employing an unconventional approach.

If your research is particularly novel or cutting-edge, any markers of broader acceptance by other, more established scholars or scholarly organs in your field can ameliorate possible concerns about its relevance and potential importance. In discussing your work, note the recognition it has received in the form of competitive grants, awards, publications in refereed journals, and/or presentations at major conferences.

It is also important to mention where you expect your research to go after the completion of your dissertation and the publications that will flow from it. They, especially at research-oriented institutions, want to know at least in broad terms where you expect to go from here. They want to see evidence of a scholarly agenda that extends beyond the dissertation. You don't need to have pages written or titles blocked out, but you need to tell them in a paragraph, (two at the most) about what questions intrigue you, and how you expect to go about finding the answers to them. These questions for future research may have been generated by the findings of your dissertation, unusual data uncovered during your fieldwork, or interesting side issues that you were forced to put off in order to keep your dissertation taut and focused. Think twice about mentioning future projects that appear entirely unrelated to your current work. Departments will want you to be firmly established in one area before you go off into another.

The relative importance of teaching versus research is a continuous and not a discrete variable. Even schools that emphasize teaching in their job listing will generally want evidence of scholarly engagement and publishing potential. It is a very competitive market for students out there, and virtually all institutions are under pressure to sell themselves via the quality of their faculty to the limited number of good students in the applicant pool. A good marker for the relative importance of teaching versus research is the teaching load. An institution with a 3-2 teaching load (a total of five courses taught per two-semester year) will expect more in the way of and value research more highly than one with a 3-3 load.

If you work in a capital-intensive area, mention your track record of gaining grants and other sources of external funding.
Teaching
In a similar fashion, the more the institution you are applying to diverges from Berkeley and the more your profile differs from the job description, the more expansive you should be in talking about your teaching and what you have to offer their students. If you have won a teaching award at Berkeley, don't make them wait until they read your CV to discover that fact.

For letters sent to large universities where you will be expected to teach large, lecture courses and graduate seminars, little is needed except to convey that you have the necessary experience and/or background. Since the dominant pedagogical style is the same as that found at Berkeley, it will be assumed that you will be able to do the same for them.

If you have TA'd some of the courses that (based on the job description) you would be expected to offer, let them know. If you have not had that opportunity, but your fields and research fall within the domain of the job description, you may want to add a line about how well prepared you are to teach such courses based on your training and research.

If your profile does not correspond exactly to the description found in the job announcement, but you believe that you could nonetheless teach the required courses, explain the basis for your confidence. What strength would you bring to the department that would more than make up for your "otherness"? Departments may decide that they don't "need" another conventional European historian if you can convince them that you can handle the core courses and offer something new that they don't currently cover as well. In this case especially, think about doing some research so you can speak in more specific terms about how you would fit into and enhance the department's offerings. Help them visualize how you would strengthen them as a department.

Small colleges will want to know that you are able to teach on a more intimate basis and are prepared to take the time to do it well. If you give them three generic lines about how "important" you consider your teaching responsibilities, don't expect much of a positive response. Spend a couple of paragraphs describing your teaching experience and philosophy, and how you would take advantage of the opportunity to create you own courses.

If you've had the opportunity to design and teach your own course, tell them, briefly, how you went about it and the choices you made. Don't just say you're a good teacher, tell them why. Look over your teaching evaluations and pick out one or two consistent strengths to highlight. Drawing on your experience, talk about how you engage students and enhance their skills (especially writing) and intellectual development.

For good, small colleges, the quality of the teaching offered to students is their stock-in- trade. Even the most research-oriented, small colleges take teaching undergraduates very seriously. You need to demonstrate that you take it seriously as well, and can talk about it using more than vague generalities and shop-worn cliches.
Closing/Enclosures
Wrap it up quickly. If you are going to be traveling for any significant portion of the job search season be sure to let them know how they can contact you (email, cellphone and/or a Pacific Bell message center account are particularly good options).

Before closing, include a sentence where you list the materials you have enclosed (e.g., CV, dissertation abstract, transcripts, teaching statement/portfolio, writing samples, etc.) and are having sent separately (letters of recommendation). You will often be assembling multiple packets at the same time (as in midnight October 14th, midnight November 14th, etc.) for jobs that ask for different combinations of enclosures. In addition to telling them what they should have received, it serves as a handy list for you to check before sealing the envelope.

Should you send a writing sample, even if they haven't asked for one? Unless the norm in your discipline is to never send more than they ask for, if you've been smart and have already run off thirty copies at two cents per page (that's only sixty cents for a thirty page chapter) by all means send it. That way it's in the file should someone become interested in you and want to read more.

Which Bioinformatics Journals Do You Follow?

Tenzin Paul — Fri, 10 Apr 2015 12:10:21 -0500

Which are your favorite bioinformatics journals? The ones that you check every month or so, or that you are subscribed to?

BINC Sample Question Paper !!!

Jitendra Narayan — Thu, 16 Apr 2015 09:15:09 -0500

BINC sample question paper round THREE ...

Reverse Complement Problem Solved with Perl

Jit — Tue, 09 Jun 2015 23:37:23 -0500

Question at http://rosalind.info/problems/1b/

#Find the reverse complement of a DNA string.
#Given: A DNA string Pattern.
#Return: Pattern, the reverse complement of Pattern.

use strict;
use warnings;

my $string="AAAACCCGGT";
my $finalString="";
my %hash = (
   "C" => "G",
   "A" => "T",
   "T" => "A",
   "G" => "C",
);

for (my $aa=0; $aa<=(length($string)-1); $aa++) {
   my $char=substr $string, $aa, 1;
   #print $hash{$char};
   $finalString="$hash{$char}"."$finalString";
}

print $finalString;
print "\n";

How long have you been a bioinformatics scientist for?

Jit — Tue, 23 Jun 2015 10:55:33 -0500

Most of the researcher have been a scientist whole life, but infact they actually started paying it with at certain time.So, how long have you been in bioinformatics domain now?