BIOL119 Lecture #3 (10/5/26)

What is the goal of sequence alignment?

maximize number of identical and similar residues in columns of alignment

What are the two steps in the alignment process after starting with unaligned sequences

1. identify alignment start

2. Insert gaps if necessary

Do the sequences themselves change during alignment?

No; the sequences do not change. Gaps are inserted to align matching residues

What four events mainly drive the evolution of biological sequences?

gene duplication

point mutations

insertions

deletions

What are the two main drivers for comparing sequences?

Functional analyses

Evolutionary analyses (divergence patterns = phylogenetics)

Why are conserved sequence regions important?

They are functionally important (similar sequences serve similar functions)

What is sequence similarity searching based on?

Alignment methods (alignment algorithms)

What are the 3 other utilities of sequence alignments (in addition to the 2 main drivers of comparing sequences)

Mutation/SNP analyses

Comparative genomics

Sequence similarity searching

What are the two main categories of sequence alignment?

pairwise sequence alignments (two sequences)

multiple sequence alignments (align many sequences together) (uses gaps?)

What are the three types of pairwise sequence alignment?

Dot plots (visualization only)

Global alignment (uses gaps?)

Local alignment (uses gaps?)

Why do we need gaps in alignments? (extra for understanding)

when sequences are biologically made, they have gene duplications, gene deletions, insertions, deletions

Point mutations include:

substitutions

insertions

deletions

Which point mutation does not require gaps and is the reason why we cannot say point mutations require gaps?

substitutions

(one residue just replaces another, so the sequence length doesn’t change)

How does a dot plot work?

Write 2 sequences into an x-y coordinate system. Identical residues (in tye sequence) get a dot. Dots form diagonals where similarities are. No diagonals = not similar

What are dot plots useful for identifying

insertions (a shift in the diagonal)

deletions (a shift in the diagonal)

and repetitive regions (multiple diagonals)

True or False: “Dot plots are an alignment method”

False. They’re a visualization - no alignment is formed

True or False: “Dots form when residues are different”

False; dots form when residues are identical

What is string matching, and how does it differ from alignment?

no gaps; (only finds exact matching regions)

What is the key difference between global and local pairwise alignment?

Global aligns the full length of both sequences using gaps

Local aligns only the most similar regions of both seq using gaps

How does multiple alignment differ from pairwise alignment? (secondary)

Multiple alignment involves more than 2 sequences. Pairwise involves exactly two.

What are the four important steps in the pairwise alignemnt process?

1. Type of alignment (global or local)

2. Scoring system

3. Algorithm to find the best scoring alignment

4. Statistics to evaluate significance

What is the first step in pairwise alignment process

Type of alignment

What is the first step in pairwise alignment process

scoring system

What is the third step in pairwise alignment process

algorithm to find best scoring alignment

What is the fourth step in pairwise alignment process

statistics to evaluate significance of alignment score

Which step is considered the core part of pairwise alignment?

Algorithm to find best scoring alignment

For divergent (dissimilar) sequences, which sequence type should you use for alignment

Protein sequences

For divergent sequences (dissimilar), why should you use protein sequences for alignment INSTEAD of DNA sequences?

Protein sequences have a 20-letter alphabet vs. DNA’s 4-letter alphabet, which gives higher information content (better scoring) and more reliable alignments

What are the three main reasons for why you should use proteins equences for alignment with divergent sequences (dissimilar)

1. better scoring system (more info)

2. reliable (less random matches)

3. functional constraints

For similar sequences, which sequence type should you use

DNA sequences

What is an easy way to understand “substitution matrix” (scoring alignment)

Basically, when you have two sequences that are aligned, we focus on two residues in a column/ We use substitution matrix to see if this alignment is good or bad

What are the three scoring parameters for alignments?

Substitution matrix

Gap opening Penalty

Gap extension penalty

What is the difference between the gap opening penalty and the gap extension penalty?

Gap opening penalty = penalty for starting a gap

Gap extension penalty = penalty for extending a gap

What does the substitution matrix score?

Residue pairs in an alignment column (matches and mismatches)

What is the substitution matrix based on

Empirically determined rates at which one residue changes into another over time, expressed as log-odds scores

(LOOKING AT HOW OFTEN RESIDUES CHANGE INTO EACH OTHER TO SEE IF WE CAN FIND A BETTER ALIGNMENT)

Substitution matrix vs. substitution

Not the same thing;

substitution matrix is used to score an alignment

substitution is a mutation (involves no gaps)

In BLOSUM matrices, what does a low number (e.g., BLOSUM50) mean vs. a high number (e.g. BLOSUM80)? (substitutions matrices)

Low number = more divergent sequences

High number = more related sequences

(Think flowers are high in number of pollen = related to me)

IN PAM matrices, what does a high number mean vs. a low number? (substitution matrices)

High number = evolutionary distant sequences

Low number = more related sequences

(think: PAM being lower = more related to me (no rizz)

What is the difference between BLOSUM and PAM in terms of the model they’re based on? (substitution matrices)

(think: numbers = function, pam = evil = evol)

BLOSUM is based on a functional model; PAM is based on an evolutionary model

For DNA and RNA alignment matrices, how are matches, mismatches, and gaps scored? (mix of substitution matrices + gap penalties)

matches = positive score

mismatches = negative score

gaps = negative gap penalty

What are the three types of gap penalties?

constant gap penalty

linear gap penalty

affine gap open and extension

Which gap penalty is most commonly used, and what two components does it have

affine gap open and extension penalties;

gap open penalty (discourages starting new gaps)

extension penalties

What is the benefit of using affine gap open and extension penalties (type of gap penalty)

gap insertions are discouraged; longer gaps are favored over many smaller short gaps

(yes discouraging of having small gaps)

What is the disadvantage of linear gap penalty? (type of gap penalty)

the overall penalty for one large gap is the same as for many small gaps that adds up to the same length

(no discouraging of having many small gaps)

What is a constant gap penalty (type of gap penalty)

every gap receives the same penalty independent of its size

What is the name of the global alignment algorithm, and who developed it?

Needleman-Wunsch algorithm. Developed by Needleman and Wunsch (1970), with an improved version Gotoh (1982).

Why can’t we compute all possible alignments between 2 sequences?

number of alignments grows exponentially — it becomes impossible to compute, even for moderately long sequences

What is the solution to the impossible number of possible alignments

dynamic programming

What is dynamic programming?

an algorithm that finds the optimal alignment much more economically

What kind of scoring system does dynamic programming use?

additive scoring system

*What are the 4 steps to aligning sequences? (more in depth)

1. type of alignment (global or local)

2. scoring system (substitution matrix + gap penalties)

3. algorithm to find best scoring alignment (dynamic programming alignment algorithm)

4. statistics to evaluate significance of alignment score

Additive scoring system (extra) does __ in order to __

just builds upon the column scores that we get from substitution matrix and gap penalties (sums the column scores together)

help us find the best scoring alignment

What are the three main steps of the dynamic programming alignment algorithm

1. Recurrence rules (popularte the matrix)

2. Boundary conditions (gaps, termination, extension).

3. Traceback (reconstruct the optimal alignment)

What is the full order to go from choosing alignment type → traceback to find optimal alignment (Global alignment)

1. Choose global alignment

2. Choose scoring methods (gap penalties + substitution matrix)

3. Set boundaries - first row and column = gap penalties

4. Fill each cell with the highest value of three options:

   1. Align two residues (substititon matrix) (corresponds to diagonal)

   2. Gap in sequence 1 (up)

   3. Gap in sequence 2 (left)

5. Additive scoring - each cell builds on the previous

6. We finish filling the matrix

7. Traceback from bottom-right: up/left = gap, diagonal = align residues

8. Final optimal alignment comes from traceback

9. Final score = bottom right cell value

What are the rules for the boundary conditions that we need in the dynamic programming alignment algorithm being used for the dynamic programming matrix (extra)

1. First row → all gap penalties (nothing from seq 1 to align)

2. First column → all gap penalties (nothing from seq 2)

3. Top left corner → 0

What are the three possible solutions at each cell in the dynamic programming matrix?

Align 2 residues and get substitution matrix score (diagonal), insert a gap in one sequence (up, gap penalty), or insert a gap in the other sequence (left, gap penalty.

Always pick whichever highest value

What does the movement in diagonal direction (up-left) signify and what score did we use?

Align 2 residues; substitution matrix

What does the movement in up direction signify and what score did we use?

gap in seq 2, gap penalty

What does the movement in left direction signify and what score did we use?

gap in sequence 1, gap penalty

Where does the traceback start, and how do you read the moves

bottom right cell. Diagonal movement = align residues. Up or left movement = insert a gap

Where is the final alignment score found?

bottom right cell of the matrix