Table of Contents

Overview and Objectives

Once raw sequence files are generated (in FASTQ format) and quality-checked, the next step in most NGS pipelines is mapping to a reference genome. For individual sequences of interest, it is common to use a tool like BLAST to identify genes or species of origin. However, a typical example NGS dataset may have tens to hundreds of millions of reads, which BLAST and similar tools are not designed to handle.

Thus, a large set of computational tools have been developed to quickly, and with sufficient (but not absolute) accuracy align each read to its best location, if any, in a reference. Even though many mapping tools exist, a few individual programs have a dominant "market share" of the NGS world. These programs vary widely in their design, inputs, outputs, and applications. In this section, we will primarily focus on two of the most versatile mappers: BWA and Bowtie2, the latter being part of the Tuxedo suite (e.g. transcriptome-aware Tophat2).

Sample Datasets

You have already worked with a paired-end yeast ChIP-seq dataset, which we will continue to use here.  The paired end data should already be located at:

$WORK/archive/original/2014_05.core_ngs

We will also use two additional RNA-seq datasets, which are located at:

$CLASSDIR/human_stuff

First copy the two human datasets to your $SCRATCH/core_ngs/fastq_prep directory.

cd $SCRATCH/core_ngs/fastq_prep
cp $CLASSDIR/human_stuff/*rnaseq.fastq.gz .

Create a $SCRATCH/core_ngs/align directory and make a link to the fastq_prep directory.

mkdir -p $SCRATCH/core_ngs/align
cd $SCRATCH/core_ngs/align
ln -s -f ../fastq_prep fq
ls -l 
ls fq

Reference Genomes

Before we get to alignment, we need a genome to align to.  We will use three different references here: 

• the human genome (hg19)
• the yeast genome (sacCer3)
• and mirbase (v20), human subset

Mirbase is a collection of all known microRNAs in all species. We will use the human subset of that database as our alignment reference.  This has the advantage of being significantly smaller than the human genome, while containing all the sequences we are actually interested in.

These are the three reference genomes we will be using today, with some information about them (and here is information about many more genomes):

Searching genomes is hard work and takes a long time if done on an un-indexed, linear genomic sequence.  So aligners require that references first be indexed for quick access  The aligners we are using each require a different index, but use the same method (the Burrows-Wheeler Transform) to get the job done. This requires taking a FASTA file as input, with each chromosome (or contig) as a separate entry, and producing some aligner-specific set of files as output. Those index files are then used by the aligner when performing the sequence alignment. 

hg19 is way too big for us to index here, so we're not going to do it. Instead, all hg19 index files are located at:

/scratch/01063/abattenh/ref_genome/bwa/bwtsw/hg19

We will grab the FASTA files for the other two references and build each index right before we use. These two references are located at:

/corral-repl/utexas/BioITeam/core_ngs_tools/references/sacCer3.fa
/corral-repl/utexas/BioITeam/core_ngs_tools/references/hairpin_cDNA_hsa.fa

First stage the yeast and mirbase reference FASTA files in your work archive area in a directory called references.

mkdir -p $WORK/archive/references/fasta
cp $CLASSDIR/references/*.fa $WORK/archive/references/fasta/

 With that, we're ready to get started on the first exercise.

Exercise #1: BWA – Yeast ChIP-seq

Overview ChIP-seq alignment workflow with BWA

We will perform a global alignment of the paired-end Yeast ChIP-seq sequences using bwa. This workflow generally has the following steps:

1. Trim the FASTQ sequences down to 50 with fastx_clipper
   • this removes most of any 5' adapter contamination without the fuss of specific adapter trimming w/cutadapt
2. Prepare the sacCer3 reference index for bwa (one time) using bwa index
3. Perform a global bwa alignment on the R1 reads (bwa aln) producing a BWA-specific binary .sai intermediate file
4. Perform a global bwa alignment on the R2 reads (bwa aln) producing a BWA-specific binary .sai intermediate file
5. Perform pairing of the separately aligned reads and report the alignments in SAM format using bwa sampe
6. Convert the SAM file to a BAM file (samtools view)
7. Sort the BAM file by genomic location (samtools sort)
8. Index the BAM file (samtools index)
9. Gather simple alignment statistics (samtools flagstat and samtools idxstat)

We're going to skip the trimming step for now and see how it goes. We'll perform steps 2 - 5 now and leave samtools for the next course section, since those steps (6 - 10) are common to nearly all post-alignment workflows.

Introducing BWA

Like other tools you've worked with so far, you first need to load bwa using the module system.  Go ahead and do that now, and then enter bwa with no arguments to view the top-level help page (many NGS tools will provide some help when called with no arguments).

module load bwa
bwa

As you can see, bwa offers a number of sub-commands one can use with to do different things.

Building the BWA sacCer3 index

We're going to index the genome with the index command. To learn what this sub-command needs in the way of options and arguments, enter bwa index with no arguments.

Usage:   bwa index [-a bwtsw|is] [-c] <in.fasta>
Options: -a STR    BWT construction algorithm: bwtsw or is [auto]
         -p STR    prefix of the index [same as fasta name]
         -6        index files named as <in.fasta>.64.* instead of <in.fasta>.*
Warning: `-a bwtsw' does not work for short genomes, while `-a is' and
         `-a div' do not work not for long genomes. Please choose `-a'
         according to the length of the genome.

Here, we only need to specify two things:

• the name of the FASTA file  
• whether to use the  bwtsw or is algorithm for indexing

Since sacCer3 is relative large (~12 Mbp) we will specify bwtsw as the indexing option, and the name of the FASTA file is sacCer3.fa.

Importantly, the output of this command is a group of files that are all required together as the index. So, within the references directory, we will create another directory called bwa/sacCer3, make a symbolic link to the yeast FASTA there, and run the index command in that directory.

mkdir -p $WORK/archive/references/bwa/sacCer3
cd $WORK/archive/references/bwa/sacCer3
ln -s ../../fasta/sacCer3.fa
ls -la

Now execute the bwa index command.

bwa index -a bwtsw sacCer3.fa

Since the yeast genome is not large when compared to human, this should not take long to execute (otherwise we would do it as a batch job). When it is comple you should see a set of index files like this:

sacCer3.fa
sacCer3.fa.amb
sacCer3.fa.ann
sacCer3.fa.bwt
sacCer3.fa.pac
sacCer3.fa.sa

Exploring the FASTA with grep

A common question is what contigs are in a given FASTA file. You'll usually want to know this before you start the alignment so that you're familiar with the contig naming convention – and to verify that it's the one you expect.

We saw that a FASTA consists of a number of contig entries, each one starting with a name line of the form below, followed by many lines of bases.

>contigName

How do we dig out just the lines that have the contig names and ignore all the sequences? Well, the contig name lines all follow the pattern above, and since the > character is not a valid base, it will never appear on a sequence line.

We've discovered a pattern (also known as a regular expression) to use in searching, and the command line tool that does regular expression matching is grep.

Regular expressions are so powerful that nearly every modern computer language includes a "regex" module of some sort. There are many online tutorials for regular expressions (and a few different flavors of them). But the most common is the Perl style (http://perldoc.perl.org/perlretut.html). We're only going to use the most simple of regular expressions here, but learning more about them will pay handsome dividends for you in the future.

Here's how to execute grep to list contig names in a FASTA file.

grep -P '^>' sacCer3.fa | more

Notes:

• The -P option tells grep to use Perl-style regular expression patterns.
  
  • This makes including special characters like Tab ( \t ), Carriage Return ( \r ) or Linefeed ( \n ) much easier that the default Posix paterns.
  • While it is not really required here, it generally doesn't hurt to include this option.

• '^>' is the regular expression describing the pattern we're looking for (described below)

• sacCer3.fa is the file to search. Lines with text that match our pattern will be written to standard output; non matching lines will be omitted.
• We pipe to more just in case there are a lot of contig names.

Now down to the nuts and bolts of our pattern, '^>'

First, the single quotes around the pattern – they are only a signal for the bash shell. As part of its friendly command line parsing and evaluation, the shell will often look for special characters on the command line that mean something to it (for example, the $ in front of an environment variable name, like in $SCRATCH). Well, regular expressions treat the $ specially too – but in a completely different way! Those single quotes tell the shell "don't look inside here for special characters – treat this as a literal string and pass it to the program". The shell will obey, will strip the single quotes off the string, and will pass the actual pattern, ^>, to the grep program. (Aside: We've see that the shell does look inside double quotes ( " ) for certain special signals, such as looking for environment variable names to evaluate.)

So what does ^> mean to grep? Well, from our contig name format description above we see that contig name lines always start with a > character, so > is a literal for grep to use in its pattern match.

We might be able to get away with just using this literal alone as our regex, specifying '>' as the command line argument. But for grep, the more specific the pattern, the better. So we constrain where the > can appear on the line. The special carat ( ^ ) character represents "beginning of line". So ^> means "beginning of a line followed by a > character, followed by anything. (Aside: the dollar sign ( $ ) character represents "end of line" in a regex. There are many other special characters, including period ( . ), question mark ( ? ), pipe ( | ), parentheses ( ( ) ), and brackets ( [ ] ), to name the most common.)

Exercise: How many contigs are there in the sacCer3 reference?

Performing the alignment

Now, we're ready to execute the actual alignment, with the goal of initially producing a SAM file from the input FASTQ files and reference. First go to the align directory, and link to the sacCer3 reference directory (this will make our commands more readable).

cd $SCRATCH/core_ngs/align
ln -s $WORK/archive/references/bwa/sacCer3
ls sacCer3

As our workflow indicated, we first use bwa aln on the R1 and R2 FASTQs, producing a BWA-specific .sai intermediate binary files. Since these alignments are completely independent, we can execute them in parallel in a batch job.

What does bwa aln needs in the way of arguments?

There are lots of options, but here is a summary of the most important ones. BWA, is a lot more complex than the options let on. If you look at the BWA manual on the web for the aln sub-command, you'll see numerous options that can increase the alignment rate (as well as decrease it), and all sorts of other things. 

The rest of the options control the details of how much a mismatch or gap is penalized, limits on the number of acceptable hits per read, and so on.  Much more information can be accessed at the BWA manual page.

For a simple alignment like this, we can just go with the default alignment parameters, with one exception. At TACC, we want to optimize our alignment speed by allocating more than one thread (-t) to the alignment. We want to run 2 tasks, and will use a minimum of one 16-core node. So we can assign 8 cores to each alignment by specifying -t 8.

Also note that bwa writes its (binary) output to standard output by default, so we need to redirect that to a .sai file. And of course we need to redirect standard error to a log file, one per file.

Create an aln.cmds file (using nano) with the following lines:

bwa aln -t 8 sacCer3/sacCer3.fa fq/Sample_Yeast_L005_R1.cat.fastq.gz > yeast_R1.sai 2> aln.yeast_R1.log
bwa aln -t 8 sacCer3/sacCer3.fa fq/Sample_Yeast_L005_R2.cat.fastq.gz > yeast_R2.sai 2> aln.yeast_R2.log

Create the batch submission script specifying a wayness of 8 (8 tasks per node) on the normal queue and a time of 1 hour, then submit the job and monitor the queue:

launcher_creator.py -n aln -j aln.cmds -t 01:00:00 -q normal -w 8
sbatch aln.slurm
showq -u

Since you have directed standard error to log files, you can use a neat trick to monitor the progress of the alignment: tail -f. The -f means "follow" the tail, so new lines at the end of the file are displayed as they are added to the file.

# Use Ctrl-c to stop the output any time
tail -f aln.yeast_R1.log

When it's done you should see two .sai files. Next we use the bwa sampe command to pair the reads and output SAM format data. For this command you provide the same reference prefix as for bwa aln, along with the two .sai files and the two original FASTQ files.

Again bwa writes its output to standard output, so redirect that to a .sam file. (Note that bwa sampe is "single threaded" – it does not have an option to use more than one processor for its work.) We'll just execute this at the command line – not in a batch job.

bwa sampe sacCer3/sacCer3.fa yeast_R1.sai yeast_R2.sai fq/Sample_Yeast_L005_R1.cat.fastq.gz fq/Sample_Yeast_L005_R2.cat.fastq.gz > yeast_pairedend.sam

You did it!  You should now have a SAM file that contains the alignments.  It's just a text file, so take a look with head, more, less, tail, or whatever you feel like.  In the next section, with samtools, you'll learn some additional ways to analyze the data once you create a BAM file.

Exercise: What kind of information is in the first lines of the SAM file?

Exercise: How many alignment records (not header records) are in the SAM file?

Exercise: How many sequences were in the R1 and R2 FASTQ files combined?

Exercises:

• Do both R1 and R2 reads have separate alignment records?
• Does the SAM file contain both aligned and un-aligned reads?
• What is the order of the alignment records in this SAM file?

Exercise #2: Bowtie2 and Local Alignment - Human microRNA-seq

Now we're going to switch over to a different aligner that was originally designed for very short reads and is frequently used for RNA-seq data. Accordingly, we have prepared another test microRNA-seq dataset for you to experiment with (not the same one you used cutadapt on). This data is derived from a human H1 embryonic stem cell (H1-hESC) small RNA dataset generated by the ENCODE Consortium – its about a half million reads.

However, there is a problem!  We don't know (or, well, you don't know) what the adapter structure or sequences were. So, you have a bunch of 36 base pair reads, but many of those reads will include extra sequence that can impede alignment – and we don't know where! We need an aligner that can find subsections of the read that do align, and discard (or "soft-clip") the rest away – an aligner with a local alignment mode. Bowtie2 is just such an aligner.

Overview miRNA alignment workflow with bowtie2

If the adapter structure were known, the normal workflow would be to first remove the adapter sequences with cutadapt. Since we can't do that, we will instead perform a local lignment of the single-end miRNA sequences using bowtie2. This workflow has the following steps:

1. Prepare the mirbase v20 reference index for bowtie2 (one time) using bowtie2-build 
2. Perform local alignment of the R1 reads with bowtie2, producing a SAM file directly
3. Convert the SAM file to a BAM file (samtools view)
4. Sort the BAM file by genomic location (samtools sort)
5. Index the BAM file (samtools index)
6. Gather simple alignment statistics (samtools flagstat and samtools idxstat)

This looks so much simpler than bwa – only one alignment step instead of three! We'll see the price for this "simplicity" in a moment...

As before, we will just do the alignment steps leave samtools for the next section.

Introducing bowtie2

Go ahead and load the bowtie2 module so we can examine some help pages and options. To do that, you must first load the perl module, and then the a specific version of bowtie2. 

module load perl
module load bowtie/2.2.0

Now that it's loaded, check out the options. There are a lot of them! In fact for the full range of options and their meaning, Google "Bowtie2 manual" and bring up that page. The Table of Contents is several pages long! Ouch!

This is the key to using bowtie2 - it allows you to control almost everything about its behavior, but that also makes it is much more challenging to use than bwa – and it's easier to screw things up too!

Building the bowtie2 mirbase index

Before the alignment, of course, we've got to build a mirbase index using bowtie2-build (go ahead and check out its options). Unlike for the aligner itself, we only need to worry about a few things here:

bowtie2-build <reference_in> <bt2_index_base>

• reference_in file is just the FASTA file containing mirbase v20 sequences
• bt2_index_base is the prefix of where we want the files to go

 Following what we did earlier for BWA indexing:

mkdir -p $WORK/archive/references/bt2/mirbase.v20
cd $WORK/archive/references/bt2/mirbase.v20
ln -s -f ../../fasta/hairpin_cDNA_hsa.fa
ls -la 

Now build the index with bowtie2-build:

bowtie2-build hairpin_cDNA_hsa.fa hairpin_cDNA_hsa.fa

That was very fast!  It's because the mirbase reference genome is so small compared to what programs like this are used to dealing with, which is the human genome (or bigger).  You should see the following files:

hairpin_cDNA_hsa.fa
hairpin_cDNA_hsa.fa.1.bt2
hairpin_cDNA_hsa.fa.2.bt2
hairpin_cDNA_hsa.fa.3.bt2
hairpin_cDNA_hsa.fa.4.bt2
hairpin_cDNA_hsa.fa.rev.1.bt2
hairpin_cDNA_hsa.fa.rev.2.bt2

Now, we're ready to actually try to do the alignment.  Remember, unlike BWA, we actually need to set some options depending on what we're after. Some of the important options for bowtie2 are:

To decide how we want to go about doing our alignment, check out the file we're aligning with less:

cd $SCRATCH/core_ngs/align
less fq/human_mirnaseq.fastq.gz

Lots of reads have long strings of A's, which must be an adapter or protocol artifact.  Even though we see how we might be able to fix it using some tools we've talked about, what if we had no idea what the adapter sequence was, or couldn't use cutadapt or other programs to prepare the reads?

 In that case, we need a local alignment where the seed length smaller than the expected insert size. Here, we are interested in finding any sections of any reads that align well to a microRNA, which are between 16 and 24 bases long, with most 20-22. So an acceptable alignment should have at least 16 matching bases, but could have more.

If we're also interested in detecting miRNA SNPs, we might want to allow a mismatch in the seed. So, a good set of options might look something like this:

-N 1 -L 16 --local

As you can tell from looking at the bowtie2 help message, the general syntax looks like this:

bowtie2 [options]* -x <bt2-idx> {-1 <m1> -2 <m2> | -U <r>} [-S <sam>]

Putting this all together and linking to the index directory, we have this. Go ahead and execute this from the command line – it will be fast.

ln -s -f $WORK/archive/reference/bt2/mirbase.v20 mb20
bowtie2 --local -N 1 -L 16 -x mb20/hairpin_cDNA_hsa.fa -U fq/human_mirnaseq.fastq.gz -S human_mirnaseq.sam

Now you should have a human_mirnaseq.sam file that you can examine using whatever commands you like. An example alignment looks like this.

TUPAC_0037_FC62EE7AAXX:2:1:2607:1430#0/1  0  hsa-mir-302b  50  22 3S20M13S * 0 0
    TACGTGCTTCCATGTTTTANTAGAAAAAAAAAAAAG  ZZFQV]Z[\IacaWc]RZIBVGSHL_b[XQQcXQcc
    AS:i:37 XN:i:0  XM:i:1  XO:i:0  XG:i:0  NM:i:1  MD:Z:16G3       YT:Z:UU

Notes:

• This is one alignment record, although it has been broken up below for readability.
• Notice the CIGAR string is 3S20M13S, meaning that 3 bases were soft clipped from one end (3S), and 13 from the other (13S).
  
  • If we did the same alignment using either bowtie2 --end-to-end mode, or using bwa aln as in Exercise #1, very little of this file would have aligned.
• The 20M part of the CIGAR string says there was a block of 20 read bases that mapped to the reference.
  • If we had not lowered the seed parameter of bowtie2 from its default of 22, we would not have found many of the alignments like this one that only matched for 20 bases.

Such is the nature of bowtie2 – it it can be a powerful tool to sift out the alignments you want from a messy dataset with limited information, but doing so requires careful tuning of the parameters, which can take quite a few trials to figure out.

Exercise #3: BWA-MEM - Human mRNA-seq

After bowtie2 came out with a local alignment option, it wasn't long before bwa developed its own local alignment algorithm called BWA-MEM (for Maximal Exact Matches), implemented by the bwa mem command. bwa mem has the following advantages:

• It incorporates a lot of the simplicity of using bwa with the complexities of local alignment, enabling straightforward alignment of datasets like the mirbase data we just examined
• It can align different portions of a read to different locations on the genome
  • In a long RNA-seq experiment, reads will (at some frequency) span a splice junction themselves, or a pair of reads in a paired-end library will fall on either side of a splice junction. We want to be able to align reads that do this for many reasons, from accurate transcript quantification to novel fusion transcript discovery.

Thus, our last exercise will be the alignment of a human long RNA-seq dataset composed (by design) almost exclusively of reads that cross splice junctions.

bwa mem was made available when we loaded the bwa module, so take a look at its usage information. The most important parameters, similar to those we've manipulated in the past two sections, are the following:

There are many more parameters to control the scoring scheme and other details, but these are the most essential ones to use to get anything of value at all.

The test file we will be working with is JUST the R1 file from a paired-end total RNA-seq experiment, meaning it is (for our purposes) single-end.  Go ahead and take a look at it, and find out how many reads are in the file.

cds
less fastq_align/human_rnaaseq.fastq.gz
gunzip -c fastq_align/human_rnaseq.fastq.gz | echo $((`wc -l`/4))

Now, try aligning it with BWA like we did in example 1.  This won't take very long, but you'll need to use our pre-indexed hg19 reference.  It is located at:

/scratch/01063/abattenh/ref_genome/bwa/bwtsw/hg19.fa

if you look at the contents of the 'bwtsw' directory in the above path, you'll see a set of files that are analogous to the yeast files we created earlier, except that their universal prefix is 'hg19.fa'.  Now, using that index, go ahead and try to do a single-end alignment of the file to the human genome like we did in Exercise #1, and save the contents to intermediate files with the prefix 'human_rnaseq_bwa'.

cds
bwa aln /scratch/01063/abattenh/ref_genome/bwa/bwtsw/hg19/hg19.fa fastq_align/human_rnaseq.fastq.gz > alignments/human_rnaseq_bwa.sai
bwa samse /scratch/01063/abattenh/ref_genome/bwa/bwtsw/hg19/hg19.fa alignments/human_rnaseq_bwa.sai fastq_align/human_rnaseq.fastq.gz > alignments/human_rnaseq_bwa.sam

Now, use less to take a look at the contents of the sam file, using the space bar to leaf through them.  You'll notice a lot of alignments look basically like this:

HWI-ST1097:228:C21WMACXX:8:1316:10989:88190     4       *       0       0       *       *       0       0
       AAATTGCTTCCTGTCCTCATCCTTCCTGTCAGCCATCTTCCTTCGTTTGATCTCAGGGAAGTTCAGGTCTTCCAGCCGCTCTTTGCCACTGATCTCCAGCT
   CCCFFFFFHHHHHIJJJJIJJJJIJJJJHJJJJJJJJJJJJJJIIIJJJIGHHIJIJIJIJHBHIJJIIHIEGHIIHGFFDDEEEDDCDDD@CDEDDDCDD

meaning they are not alignments at all.  Essentially, nothing (with a few exceptions) aligned.  That's because this file was generated exclusively from reads in a larger dataset that cross at least one splice junction (surprise!), meaning that they sequence as it exists in most of the reads does not exist anywhere in the genome, but some subsections of each read do exist somewhere in the genome.  So, we need an aligner that is capable of finding regions of each read (above some length cutoff) that align to the genome.  Based on the following syntax, and using the reference path, we noted earlier, try to use BWA MEM to align the same file, single-end, saving all output files with the prefix 'human_rnaseq_mem':

bwa mem ref.fa reads.fq > aln-se.sam

bwa mem /scratch/01063/abattenh/ref_genome/bwa/bwtsw/hg19/hg19.fa fastq_align/human_rnaseq.fastq.gz > alignments/human_rnaseq_mem.sam

Now, check the length of the SAM file you generated with 'wc -l'.  Since there is one alignment per line, there must be 586266 alignments, which is more than the number of sequences in the FASTQ file.  This is because many reads will align twice or more, hopefully on either side of a splice junction.  These alignments can still be associated because they will have the same read ID, but are reflected in more than one line.  To get an idea of how often each read aligned, and what the 'real' alignment rate is, use the following commands:

cut -f 1 alignments/human_rnaseq_mem.sam | sort | uniq -c | less	#This gives you a view where each read is listed next to the number of entries it has in the SAM file
cut -f 1 alignments/human_rnaseq.sam | sort | uniq -c | awk '{print $1}' | sort | uniq -c | less	#This gives essentially a histogram of the number of times each read aligned - a plurality of reads aligned twice, which seems reasonable since these are all reads crossing a junction, but plenty aligned more or less
cut -f 1 alignments/human_rnaseq.sam | sort | uniq | wc -l	#This gives a better idea of the alignment rate, which is how many reads aligned at least once to the genome.  Divided by the number of reads in the original file, the real alignment rate is around 64.19 %.

# PLEASE NOTE: some of these one-liners are only reasonably fast if the files are relatively small (around a million reads or less).  For bigger files, there are better ways to get this information, mostly using samtools, which is a utility explicitly designed for manipulating SAM/BAM files.  We'll cover samtools in the next section.

This alignment rate is pretty good, but it could get better by playing around with the finer details of bwa mem.

BWA-MEM vs Tophat

Another approach to aligning long RNA-seq data is to use an aligner that is more explicitly concerned with sensitivity to splice sites, namely a program like Tophat. Tophat uses either bowtie (tophat) or bowtie2 (tophat2) as the actual aligner, but performs the following steps:

• aligns reads to the genome
• reads that do not align to the genome are aligned against a transcriptome, if provided
  • if they align, the transcriptome coordinates are converted back to genomic coordinates, with gaps represented in the CIGAR string, for example as 196N
• reads that do not align to the transcriptome are split into smaller pieces, each of which Tophat attempts to map to the genome

Note that Tophat also reports secondary alignments, but they have a different meaning. Tophat always reports spliced alignments as one alignment records with the N CIGAR string operator indicating the gaps. Secondary alignments for Tophat (marked with the 0x100 BAM flag) represent alternate places in the genome where a  read (spliced or not) may have mapped.

As you can imagine from this series of steps, Tophat is very computationally intensive and takes much longer than bwa mem – very large alignments (hundreds of millions of reads) may not complete in stampede's 48 hour maximum job time!