Bioinformatics

Build tries, Aho-Corasick, and suffix arrays with Kasai LCP to index DNA/text and match many patterns in one pass. Use for genome motif scanning, k-mer indexing, longest-repeat search, or BWA/FM-index groundwork.

Interpret AlphaFold2/AF3 pLDDT/PAE scores, fetch AlphaFold DB models by UniProt ID, and rank RFdiffusion/ProteinMPNN designs. Use when asked about pLDDT, PAE, AF2 vs AF3, AlphaFold DB fetch, or design triage.

Code DDPM/DDIM diffusion samplers, linear/cosine noise schedules, and DDRM inverse-problem solving (denoising, inpainting, super-resolution) in NumPy/PyTorch. Use for forward/reverse diffusion, score matching, or DDIM sampling.

Predict CAGE/DNase/ATAC/ChIP-seq tracks from raw DNA with Enformer/Borzoi, run in-silico mutagenesis (ISM), and score noncoding variant effects. Use when predicting enhancer/promoter activity from sequence, running ISM, scoring a noncoding SNP, or prioritizing GWAS/eQTL variants.

Choose Borzoi (RNA-seq coverage, 32bp) vs Epiformer (sequence+PhyloP, chromatin accessibility) vs AlphaGenome for epigenomic prediction. Use when picking a model for RNA-seq, ATAC/DNase, or variant-effect scoring.

Generate ESM2 protein embeddings (fair-esm/transformers) and predict structure with ESMFold. Use when embedding sequences, scoring mutations zero-shot, annotating protein function, or doing fast MSA-free structure prediction.

Tokenize scRNA-seq via Geneformer gene-rank or scGPT expression-bin encoding; annotate cell types, simulate in-silico knockouts. Use for foundation-model cell annotation, Geneformer/scGPT tokenization, or perturbation prediction.

Embed DNA with genomic foundation models (Nucleotide Transformer, HyenaDNA, Evo) via HuggingFace transformers; k-mer tokenize, probe promoter motifs. Use for DNA LLMs, genomic embeddings, or NT/HyenaDNA/Evo choice.

Fine-tune LLMs (Mistral/Llama/Qwen) with LoRA/QLoRA via HuggingFace PEFT, bitsandbytes NF4, and trl SFTTrainer. Use when doing LoRA/QLoRA fine-tuning, PEFT, instruction tuning, or SFTTrainer on limited GPU memory.

Track LLM fine-tuning runs with pandas/dataclasses + W&B/MLflow/TensorBoard: log per-epoch loss, run one-factor ablations (LoRA rank, LR), build a run registry. Use for hyperparameter ablations or picking an early-stop epoch.

Score splicing variant effects with SpliceAI/Pangolin delta scores (DS_AG/DS_AL/DS_DG/DS_DL) and AlphaGenome. Use when scoring a VCF for splice disruption, interpreting DS thresholds, or ranking cryptic splice-site variants.

Triage variant scores for structural follow-up; pick AlphaFold2 vs AlphaFold3 vs RoseTTAFold2 and gate results on pLDDT/PAE. Use for monomer/complex/ligand model choice or ranking variants by structural confidence.

ColPali-style Vision RAG: embed rendered PDF pages, retrieve via ColBERT MaxSim, feed top-k pages to Qwen2-VL, no OCR. Use for PDF/document QA over figures and tables, multimodal retrieval, or Recall@k/MRR eval.

Score protein point mutations zero-shot with ESM-1v/ESM-2 masked-LM log-odds, ensembled, benchmarked on ProteinGym DMS. Use when predicting mutation effects, ranking missense variants, scoring VUS fitness with no labels.

Build an Aho-Corasick automaton (trie + BFS failure links) to find every pattern occurrence in one O(n+m+z) text pass. Use for restriction-site/primer/motif search or replacing per-pattern KMP/regex loops over many fixed patterns.

Implement a self-balancing AVL binary search tree in Python with rotation-based rebalancing (LL/RR/LR/RL) guaranteeing O(log n) insert/delete/search. Use when a user asks to build/implement an AVL tree, keep a sorted index balanced under insert/delete, explain balance factor or tree rotations, or avoid O(n) degeneration of a BST on sorted/near-sorted input (e.g. genomic positions arriving in coordinate order).

Compute GCD/LCM via Euclid's algorithm, find roots with Newton's method, and count k-mers using dict/Counter vs O(n^2) list-scan. Use for GCD/LCM math, root-finding, k-mer counting, or Big-O complexity questions.

BFS (queue) and DFS (stack/recursion) traversal in pure Python for shortest unweighted path, k-hop neighborhood, connected components, cycle detection on PPI/regulatory networks. Use for shortest path or cycle detection.

Implement/debug a binary search tree in Python: insert, search, delete (3 cases), successor/predecessor, inorder/level-order traversal. Use for BST coding, O(log n) vs O(n) degenerate cases, or choosing AVL/Red-Black.

Implement bubble/merge/shell/quicksort in Python; compare Big-O time/space/stability. Use when asked to sort an array, code a sort from scratch, explain quicksort complexity, or fix O(n^2) worst case on sorted input.

Derive Big O time/space complexity of loops and recursion via recurrence relations; simplify expressions, compare growth at scale. Use when asked the complexity of code, hunting O(n^2) loops, or worst-case cost.

Build a DFA transition table via the KMP prefix function, then scan text in O(n) with zero backtracking. Use when repeatedly searching one fixed pattern (motif, restriction site, primer) against many sequences or texts.

Compute single-source shortest paths in a non-negative-weight graph with Dijkstra's algorithm (binary-heap priority queue, O((V+E) log V)); reconstruct paths and find network diameter. Use when finding shortest/cheapest/most-reliable path, routing, weighted PPI/interaction-network distance, or ranking paths by confidence score product.

Implement resizable arrays with ctypes-backed doubling/shrinking; prove append is amortized O(1). Use when asked why list.append is O(1), to build a DynamicArray class, or to compare list vs numpy append speed.

Build graph structures (adjacency matrix/list, edge list) in Python/NumPy for PPI, GRN, and metabolic networks. Use when representing a graph, picking sparse vs dense storage, loading an edge-list file, or prepping for BFS/DFS/Dijkstra/MST.

Implement Python hash tables (chaining, open addressing, rehashing) and Bloom filters for set membership. Use when building a hash table from scratch, resolving hash collisions, sizing a Bloom filter, or checking k-mer/key set membership under memory limits.

Speed up exponential recursion (Fibonacci, alignment counting, coin change) to linear time via dict cache or lru_cache. Use when recursion is slow, or asked to memoize, add @lru_cache, or explain overlapping subproblems.

Find all overlapping exact occurrences of a pattern/motif/primer in a string or DNA sequence in O(n+m) time via KMP's prefix/failure-function. Use for exact substring search, motif/primer location, or a slow naive O(n*m) scan.

Solve 0/1, unbounded, subset-sum, and bitmask set-cover knapsack DP in Python with traceback and O(capacity)-space optimization. Use when picking an optimal subset under a budget/capacity constraint — gene panel or assay selection under a sequencing budget, primer/reagent allocation, experiment portfolio selection, or any "maximize value subject to a cost limit" or "does a subset sum to X" problem.

Implement Python linear/binary search: first/last occurrence, lower_bound/upper_bound (bisect), rotated-sorted-array search. Use when finding an index, searching sorted data, counting duplicates, or finding an insertion point.

Implement counting sort, radix sort, and bucket sort in Python for O(n) non-comparison sorting of integers, fixed-length strings, and DNA k-mers. Use when sorting integers with a small known range, sorting fixed-length keys/k-mers for de Bruijn graph construction or k-mer analysis, or explaining why non-comparison sorts beat the Omega(n log n) lower bound.

Implement singly/doubly linked lists in Python (O(1) head/tail insert, delete, reverse) plus pointer problems like Floyd's cycle detection and merge-sorted-lists. Use for linked-list coding-interview questions.

Compute minimum spanning trees with Kruskal's (Union-Find) and Prim's (min-heap) algorithms in Python or networkx. Use when building a phylogenetic distance tree, gene co-expression network backbone, MST-based clustering, or implementing Union-Find/disjoint-set.

Brute-force O(n*m) sliding-window search for all overlapping matches of a pattern/motif/primer in text or DNA/protein strings, pure Python. Use for one-off exact search, or to benchmark the naive baseline before KMP/Rabin-Karp/Boyer-Moore.

Rabin-Karp rolling-hash search in Python for single/multi-pattern matching (DNA motifs, k-mers, plagiarism phrases). Use when finding pattern occurrences in text, explaining rolling hash, or comparing vs KMP/naive search.

Implement a red-black self-balancing BST (insert, rotations, recoloring) for O(log n) search on sorted VCF variant positions. Use when building a balanced BST, verifying invariants, or comparing red-black vs AVL trees.

Implement Needleman-Wunsch global and Smith-Waterman local sequence alignment: fill/traceback DP matrices, match/mismatch or BLOSUM62 scoring. Use when coding alignment from scratch or explaining DP traceback algorithms.

Implement Stack (LIFO)/Queue (FIFO) in Python (array, linked-list, two-stack) with O(1) ops; validate balanced brackets/RNA dot-bracket notation. Use for stack/queue from scratch, backing BFS/DFS, or checking parens.

Build a suffix array (Manber-Myers O(n log n)) and LCP array (Kasai's O(n)) in Python; binary-search substrings, count k-mers, find longest repeated motifs. Use for text indexing, pattern search, or aligner (BWA-like) internals.

Build a suffix tree for O(m) pattern search, longest repeated substring, and longest common substring (LCS). Use when finding all motif occurrences in DNA/text, detecting tandem repeats, or comparing two sequences' shared region.