Update CSE5313_L26.md
This commit is contained in:
@@ -33,7 +33,7 @@ mRNA carries info to Ribosome as codons of length 3 over GUCA.
|
|||||||
|
|
||||||
2nd Chargaff rule:
|
2nd Chargaff rule:
|
||||||
|
|
||||||
- $#𝐴 ≈ #𝑇$ and $#G \approx #C$ in each strands.
|
- $#A \approx #T$ and $#G \approx #C$ in each strands.
|
||||||
- Can be explained via tandem duplications.
|
- Can be explained via tandem duplications.
|
||||||
- $GCAGCATT \implies GCAGCAGCATT$.
|
- $GCAGCATT \implies GCAGCAGCATT$.
|
||||||
- Occur naturally during cell mitosis.
|
- Occur naturally during cell mitosis.
|
||||||
@@ -369,30 +369,31 @@ This implies that $n-\log |\mathcal{C}|\geq n-\log(n+1)-\log|\mathcal{C}_j|\geq
|
|||||||
|
|
||||||
Corollary: In the relevant regime $t=O(n^{1-\epsilon})$, we have $\Omega(t\log n)$ redundancy.
|
Corollary: In the relevant regime $t=O(n^{1-\epsilon})$, we have $\Omega(t\log n)$ redundancy.
|
||||||
|
|
||||||
TRACK LOST HERE
|
### t-break codes: Main ideas.
|
||||||
|
|
||||||
𝑡-break codes: Main ideas.
|
Encoding:
|
||||||
• Encoding:
|
|
||||||
– Need multiple markers across the codeword.
|
|
||||||
– Construct an adjacency matrix 𝐴 of markers to record their order.
|
|
||||||
– Append 𝑅𝑆2𝑡 𝐴 to the codeword (as in the sliced channel).
|
|
||||||
• Decoding (from 𝑡 + 1 fragments):
|
|
||||||
– Locate all surviving markers, and locate 𝑅𝑆2𝑡 𝐴 ′.
|
|
||||||
– Build an approximate adjacency matrix 𝐴
|
|
||||||
′
|
|
||||||
from surviving markers (𝑑𝐻 𝐴, 𝐴′ ≤ 2𝑡).
|
|
||||||
– Correct 𝐴
|
|
||||||
′
|
|
||||||
, 𝑅𝑆2𝑡 𝐴 ′ ↦ 𝐴 , 𝑅𝑆2𝑡 𝐴 .
|
|
||||||
– Order the fragments correctly using 𝐴.
|
|
||||||
• Tools:
|
|
||||||
– Random encoding (to have many markers).
|
|
||||||
– Mutually uncorrelated codes (so that markers will not overlap).
|
|
||||||
|
|
||||||
Tool: Mutually uncorrelated codes.
|
- Need multiple markers across the codeword.
|
||||||
• Want: Markers not to overlap.
|
- Construct an adjacency matrix 𝐴 of markers to record their order.
|
||||||
• Solution: Take markers from a Mutually Uncorrelated Codes (existing notion).
|
- Append $RS_{2t}(A)$ to the codeword (as in the sliced channel).
|
||||||
– A code ℳ is called mutually uncorrelated if no suffix of any 𝑚𝑖 ∈ ℳ if a prefix of another
|
|
||||||
|
Decoding (from $t + 1$ fragments):
|
||||||
|
|
||||||
|
- Locate all surviving markers, and locate $RS_{2t}(A)'$.
|
||||||
|
- Build an approximate adjacency matrix $A'$ from surviving markers $(d_H(A, A' )\leq 2t)$.
|
||||||
|
- Correct $(A',RS_{2t}(A)')\mapsto (A,RS_{2t}(A))$.
|
||||||
|
- Order the fragments correctly using $A$.
|
||||||
|
|
||||||
|
Tools:
|
||||||
|
|
||||||
|
- Random encoding (to have many markers).
|
||||||
|
- Mutually uncorrelated codes (so that markers will not overlap).
|
||||||
|
|
||||||
|
#### Tool: Mutually uncorrelated codes.
|
||||||
|
|
||||||
|
- Want: Markers not to overlap.
|
||||||
|
- Solution: Take markers from a Mutually Uncorrelated Codes (existing notion).
|
||||||
|
- A code $\mathcal{M}$ is called mutually uncorrelated if no suffix of any 𝑚𝑖 ∈ ℳ if a prefix of another
|
||||||
𝑚𝑗 ∈ ℳ (including 𝑖 = 𝑗).
|
𝑚𝑗 ∈ ℳ (including 𝑖 = 𝑗).
|
||||||
– Many constructions exist.
|
– Many constructions exist.
|
||||||
• Theorem: For any integer ℓ there exists a mutually uncorrelated code 𝐶𝑀𝑈 of length
|
• Theorem: For any integer ℓ there exists a mutually uncorrelated code 𝐶𝑀𝑈 of length
|
||||||
|
|||||||
Reference in New Issue
Block a user